P. 1
Handbook of Soil Analysis

Handbook of Soil Analysis

4.5

|Views: 365|Likes:
Published by api-3699380

More info:

Published by: api-3699380 on Oct 17, 2008
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

03/18/2014

pdf

text

original

Marc Pansu Jacques Gautheyrou Handbook of Soil Analysis Mineralogical, Organic and Inorganic Methods

Marc Pansu Jacques Gautheyrou

Handbook of Soil Analysis
Mineralogical, Organic and Inorganic Methods
with 183 Figures and 84 Tables

Dr Marc Pansu Centre IRD BP 64501 Avenue Agropolis 911 34394 Montpellier Cedex 5 France
E-mail : pansu@mpl.ird.fr

Jacques Gautheyrou Avenue de Marinville 6 94100 St. Maur des Fossés France

Updated English version, corrected by Daphne Goodfellow. The original French book "L'analyse du sol, minéralogique et minérale" by Marc Pansu and Jacques Gautheyrou, was published in 2003 by Springer-Verlag , Berlin Heidelberg New York.

Library of Congress Control Number: 2005938390 ISBN-10 3-540-31210-2 Springer Berlin Heidelberg New York ISBN-13 978-3-540-31210-9 Springer Berlin Heidelberg New York
This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer-Verlag. Violations are liable to prosecution under the German Copyright Law. Springer is a part of Springer Science+Business Media springer.com © Springer-Verlag Berlin Heidelberg 2006 Printed in The Netherlands The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Cover design: E. Kirchner, Heidelberg Production: Almas Schimmel Typesetting: SPI Publisher Services Printing: Krips bv, Meppel Binding: Stürtz AG, Würzburg Printed on acid-free paper 30/3141/as 5 4 3 2 1 0

FOREWORD
This new book by Marc Pansu and Jacques Gautheyrou provides a synopsis of the analytical procedures for the physicochemical analysis of soils. It is written to conform to analytical standards and quality control. It focuses on mineralogical, organic and inorganic analyses, but also describes physical methods when these are a precondition for analysis. It will help a range of different users to choose the most appropriate method for the type of material and the particular problems they have to face. The compiled work is the product of the experience gained by the authors in the laboratories of the Institute of Research for Development (IRD) in France and in tropical countries, and includes an extensive review of the literature. The reference section at the end of each chapter lists source data from pioneer studies right up to current works, such as, proposals for structural models of humic molecules, and itself represents a valuable source of information. IRD soil scientists collected data on Mediterranean and tropical soils in the field from West and North Africa, Madagascar, Latin America, and South East Asia. Soil materials from these regions are often different from those found in temperate zones. As their analysis brought new problems to light, it was essential to develop powerful and specific physicochemical methods. Physicists, chemists and biologists joined forces with IRD soil scientists to contribute knowledge from their own disciplines thereby widening its scope considerably. This work is the fruit of these experiments as applied to complex systems, involving soils and the environment. The methodological range is particularly wide and each chapter presents both simple analyses and analyses that may require sophisticated equipment, as well as specific skills. It is aimed both at teams involved in practical field work and at researchers involved in fundamental and applied research. It describes the principles, the physical and chemical basis of each method, the corresponding analytical procedures, and the constraints and limits of each. The descriptions are practical, easy to understand and implement. Summary tables enable a rapid overview of the data. Complex techniques are explained under the heading ‘Principle’ and concrete examples of methods include: spectra (near and far IR, UVvisible, 1H-NMR, 13C-NMR, ESR, ICP-AES, ICP-MS, X-ray fluorescence, EDX or WDX microprobe, neutron activation analysis), diffractograms (XRD, electron microdiffraction), thermograms (DTA, DTG, TGA), chromatograms (GPC, HPLC, ionic chromatography, exclusion chromatography), electrophoregrams, ion exchange methods, electrochemistry, biology, different physical separation techniques, selective dissolutions, and imagery.

VI

Foreword

The book will be valuable not only for researchers, engineers, technicians and students in soil science, but also for agronomists and ecologists and others in related disciplines, such as, analytical physical chemistry, geology, climatology, civil engineering and industries associated with soil. It is a basic work whose goal is to contribute to the scientific analysis of the environment. The methodologies it describes apply to a wide range of bioclimatic zones: temperate, arid, subtropical and tropical. As with the previous books by the same authors (Pansu, Gautheyrou and Loyer, 1998, Masson, Paris, Milan, Barcelona; Pansu, Gautheyrou and Loyer, 2001, Balkema, Lisse, Abington, Exton, Tokyo), this new book represents a reference work for our laboratories. We are confident its originality and ease of use will ensure its success. Alain Aventurier, Director of Analytical Laboratories of CIRAD1 Christian Feller, Director of Research at IRD 2 Pierre Bottner, Director of Research at CNRS 3

1

2 3

CIRAD, Centre International pour la Recherche Agronomique et le Développement (France). IRD, Institut de Recherche pour le Développement (ex ORSTOM, France). CNRS, Centre National de la Recherche Scientifique (France).

CONTENTS
PART 1 - MINERALOGICAL ANALYSIS
CHAPTER 1 Water Content and Loss on Ignition
1.1 Introduction ..................................................................................................3 1.2 Water Content at 105°C (H2O−) ....................................................................6 1.2.1 Principle .................................................................................................6 1.2.2 Materials ................................................................................................6 1.2.3 Sample...................................................................................................6 1.2.4 Procedure ..............................................................................................7 1.2.5 Remarks ................................................................................................7 + 1.3 Loss on Ignition at 1,000°C (H2O ) ..............................................................8 1.3.1 Introduction ............................................................................................8 1.3.2 Principle ...............................................................................................11 1.3.3 Equipment............................................................................................11 1.3.4 Procedure ............................................................................................11 1.3.5 Calculations .........................................................................................12 1.3.6 Remarks ..............................................................................................12 Bibliography .....................................................................................................12

CHAPTER 2 Particle Size Analysis
2.1 Introduction ................................................................................................15 2.1.1 Particle Size in Soil Science ................................................................15 2.1.2 Principle ...............................................................................................17 2.1.3 Law of Sedimentation ..........................................................................18 2.1.4 Conditions for Application of Stokes Law.............................................24 2.2 Standard Methods ......................................................................................26 2.2.1 Pretreatment of the Sample .................................................................26 2.2.2 Particle Suspension and Dispersion ....................................................31 2.2.3 Pipette Method after Robinson-Köhn or Andreasen ............................35 2.2.4 Density Method with Variable Depth ....................................................42 2.2.5 Density Method with Constant Depth...................................................47 2.2.6 Particle Size Analysis of Sands Only ...................................................48 2.3 Automated Equipment ...............................................................................50 2.3.1 Introduction ..........................................................................................50 2.3.2 Method Using Sedimentation by Simple Gravity..................................51 2.3.3 Methods Using Accelerated Sedimentation .........................................53 2.3.4 Methods Using Laser Scattering and Diffraction..................................54 2.3.5 Methods Using Optical and Electric Properties....................................55 2.3.6 Methods Allowing Direct Observations of the Particles........................55 2.3.7 Methods Using Conductivity ................................................................56 References ........................................................................................................56 Bibliography .....................................................................................................58 Generality .....................................................................................................58

VIII

Contents

Pre-treatment................................................................................................58 Pipette Method..............................................................................................61 Hydrometer Method ......................................................................................62 Instrumental Methods ...................................................................................62

CHAPTER 3 Fractionation of the Colloidal Systems
3.1 Introduction ................................................................................................ 65 3.2 Fractionation by Continuous Centrifugation........................................... 66 3.2.1 Principle............................................................................................... 66 3.2.2 Theory ................................................................................................. 69 3.2.3 Equipment and reagents ..................................................................... 73 3.2.4 Procedure............................................................................................ 75 3.3 Pretreatment of the Extracted Phases ..................................................... 79 References........................................................................................................ 81 Bibliography ..................................................................................................... 81

CHAPTER 4 Mineralogical Characterisations by X-Ray Diffractometry
4.1 Introduction ................................................................................................ 83 4.1.1 X-Ray Diffraction and Mineralogy ........................................................ 83 4.1.2 Principle............................................................................................... 86 4.1.3 XRD Instrumentation ........................................................................... 87 4.2 Qualitative Diffractometry.......................................................................... 90 4.2.1 Overview of Preparation of the Samples ............................................. 90 4.2.2 Preparation for Powder Diagrams ....................................................... 90 4.2.3 Preparation for Oriented Diagrams...................................................... 94 4.2.4 Pretreatment of Clays.......................................................................... 99 4.2.5 Qualitative Diffractometry ..................................................................113 4.3 Quantitative Mineralogical Analysis .......................................................118 4.3.1 Interest ..............................................................................................118 4.3.2 Quantitative Mineralogical Analysis by XRD......................................118 4.3.3 Multi-Instrumental Quantitative Mineralogical Analysis......................124 References......................................................................................................126 Bibliography ...................................................................................................127 General.......................................................................................................127 Preparation of Oriented Aggregates on Porous Ceramic Plate ..................128 Saturation of Clays by Cations ...................................................................129 Saturation, Solvation, Intercalation Complex, Dissolution ..........................129 Preparation of Iron Oxides..........................................................................130 Quantitative XRD........................................................................................130

CHAPTER 5 Mineralogical Analysis by Infra-Red Spectrometry
5.1 Introduction ..............................................................................................133 5.1.1 Principle.............................................................................................133 5.1.2 IR Instrumentation .............................................................................135 5.2 IR Spectrometry in Mineralogy................................................................138 5.2.1 Equipment and Products ...................................................................138 5.2.2 Preparation of the Samples ...............................................................139 5.2.3 Brief Guide to Interpretation of the Spectra....................................... 146 5.2.4 Quantitative Analysis .........................................................................152

Contents

IX

5.3 Other IR Techniques ................................................................................156 5.3.1 Near-infrared Spectrometry (NIRS)................................................... 156 5.3.2 Coupling Thermal Measurements and FTIR Spectrometry of Volatile Products ............................................................................................158 5.3.3 Infrared Microscopy ...........................................................................159 5.3.4 Raman Scattering Spectroscopy ...................................................... 159 References......................................................................................................161 Chronobibliography.......................................................................................162

CHAPTER 6 Mineralogical Separation by Selective Dissolution
6.1 Introduction ............................................................................................. 167 6.1.1 Crystallinity of Clay Minerals............................................................. 167 6.1.2 Instrumental and Chemical Methods ................................................ 169 6.1.3 Selective Dissolution Methods .......................................................... 172 6.1.4 Reagents and Synthetic Standards .................................................. 174 6.2 Main Selective Dissolution Methods...................................................... 180 6.2.1 Acid Oxalate Method Under Darkness (AOD)................................... 180 6.2.2 Dithionite-Citrate-Bicarbonate Method (DCB) ................................... 187 6.2.3 EDTA Method ................................................................................... 192 6.2.4 Pyrophosphate Method..................................................................... 196 6.2.5 Extraction in Strongly Alkaline Mediums ........................................... 201 6.3 Other Methods, Improvements and Choices ........................................ 206 6.3.1 Differential Sequential Methods ........................................................ 206 6.3.2 Selective Methods for Amorphous Products ..................................... 210 6.3.3 Brief Overview to the Use of the Differential Methods ...................... 214 References ..................................................................................................... 215

CHAPTER 7 Thermal Analysis
7.1 Introduction ............................................................................................. 221 7.1.1 Definition........................................................................................... 221 7.1.2 Interest.............................................................................................. 223 7.2 Classical Methods ................................................................................... 226 7.2.1 Thermogravimetric Analysis.............................................................. 226 7.2.2 Differential Thermal Analysis and Differential Scanning Calorimetry 235 7.3 Multi-component Apparatuses for Thermal Analysis........................... 246 7.3.1 Concepts........................................................................................... 246 7.3.2 Coupling Thermal Analysis and Evolved Gas Analysis..................... 247 References ..................................................................................................... 249 Chronobibliography ...................................................................................... 250

CHAPTER 8 Microscopic Analysis
8.1 Introduction ............................................................................................. 253 8.2 Preparation of the Samples .................................................................... 254 8.2.1 Interest.............................................................................................. 254 8.2.2 Coating and Impregnation, Thin Sections ......................................... 255 8.2.3 Grids and Replicas for Transmission Electron Microscopy ............... 261 8.2.4 Mounting the Samples for Scanning Electron Microscopy ................ 263 8.2.5 Surface Treatment (Shadowing, Flash-carbon, Metallization) .......... 265

X

Contents

8.3 Microscope Studies................................................................................. 267 8.3.1 Optical Microscopy ........................................................................... 267 8.3.2 Electron Microscopy, General Information ........................................ 270 8.3.3 Transmission Electron Microscopy, Micro-diffraction ........................ 271 8.3.4 Scanning Electron Microscopy.......................................................... 279 8.3.5 Ultimate Micro-analysis by X-Ray Spectrometry............................... 282 References ..................................................................................................... 283 Chronobibliography ...................................................................................... 284

PART 2 - ORGANIC ANALYSIS
CHAPTER 9 Physical Fractionation of Organic Matter
9.1 Principle and Limitations ........................................................................289 9.1.1 Forms of Organic Matter in Soil .........................................................289 9.1.2 Principle.............................................................................................289 9.1.3 Difficulties ..........................................................................................291 9.2 Methods ....................................................................................................293 9.2.1 Classification .....................................................................................293 9.2.2 Extraction of Plant Roots ...................................................................293 9.2.3 Dispersion of the Particles.................................................................296 9.2.4 Separation by Density. ......................................................................309 9.2.5 Particle Size Fractionations ...............................................................314 9.2.6 Precision of the Fractionation Methods .............................................320 9.3 Conclusion and Outlook..........................................................................321 References......................................................................................................322

CHAPTER 10 Organic and Total C, N (H, O, S) Analysis
10.1 Introduction ............................................................................................327 10.1.1 Soil Organic Matter..........................................................................327 10.1.2 Sampling, Preparation of the Samples, Analytical Significance.......330 10.2 Wet Methods...........................................................................................333 10.2.1 Total Carbon: General Information ..................................................333 10.2.2 Organic Carbon by Wet Oxidation at the Temperature of Reaction ......................................................................................335 10.2.3 Organic Carbon by Wet Oxidation at Controlled Temperature ........340 10.2.4 Organic Carbon by Wet Oxidation and Spectrocolorimetry..............342 10.2.5 Total Nitrogen by Wet Method: Introduction ........ ............................342 10.2.6 Total Nitrogen by Kjeldahl Method and Titrimetry ...........................344 10.2.7 Kjeldahl N, Titration by Spectrocolorimetry......................................349 10.2.8 Kjeldahl N, Titration by Selective Electrode . ....... ............................351 10.2.9 Mechanization and Automation of the Kjeldahl Method...................353 10.2.10 Modified Procedures for NO3–, NO2– and Fixed N .........................354 10.3 Dry Methods ...........................................................................................355 10.3.1 Total Carbon by Simple Volatilization ..............................................355 10.3.2 Simultaneous Instrumental Analysis by Dry Combustion: CHN(OS)356 10.3.3 CHNOS by Thermal Analysis ..........................................................362

Contents

XI

10.3.4 C and N Non-Destructive Instrumental Analysis..............................363 10.3.5 Simultaneous Analysis of the Different C and N Isotopes ...............364 References......................................................................................................365 Bibliography ...................................................................................................367

CHAPTER 11 Quantification of Humic Compounds
11.1 Humus in Soils .......................................................................................371 11.1.1 Definitions........................................................................................371 11.1.2 Role in the Soil and Environment ....................................................373 11.1.3 Extractions.......................................................................................374 11.2 Main Techniques ....................................................................................375 11.2.1 Extraction ........................................................................................375 11.2.2 Quantification of the Extracts...........................................................379 11.2.3 Precision and Correspondence of the Extraction Methods ............. 383 11.2.4 Purification of Humic Materials ....................................................... 389 11.3 Further Alternatives and Complements Methods............................... 392 11.3.1 Alternative Method of Extraction ..................................................... 392 11.3.2 Fractionation of the Humin Residue................................................ 392 References ..................................................................................................... 395 Humic Materials ......................................................................................... 395 Extraction, Titration, Purification and Fractionation of Humic Materials ..... 396

CHAPTER 12 Characterization of Humic Compounds
12.1 Introduction ........................................................................................... 399 12.1.1 Mechanisms of Formation............................................................... 399 12.1.2 Molecular Structure......................................................................... 400 12.2. Classical Techniques ........................................................................... 401 12.2.1 Fractionation of Humic Compounds................................................ 401 12.2.2 Titration of the Main Functional Groups .......................................... 408 12.2.3 UV–Visible Spectrometry ................................................................ 410 12.2.4 Infra-Red Spectrography................................................................. 413 12.3 Complementary Techniques ................................................................ 415 12.3.1 Improvements in Fractionation Technologies ................................. 415 12.3.2 Titration of Functional Groups......................................................... 418 12.3.3 Characterization by Fragmentation ................................................. 419 12.3.4 Nuclear Magnetic Resonance (NMR) ............................................. 424 12.3.5 Fluorescence Spectroscopy............................................................ 433 12.3.6 Electron Spin Resonance (ESR) Spectroscopy .............................. 435 12.3.7 Measurement of Molecular Weight and Molecular Size ................. 437 12.3.8 Microscopic Observations............................................................... 440 12.3.9 Other Techniques ........................................................................... 441 References ..................................................................................................... 442 Molecular Models....................................................................................... 442 Fractionation, Determination of Molecular Weights and Molecular Sizes .. 443 Functional Group of Humic Compounds .................................................... 445 Spectrometric Characterizations................................................................ 446 UV–Visible, IR, Fluorescence, ESR Spectrometries .................................. 446 Nuclear Magnetic Resonance.................................................................... 447

XII

Contents

Methods of Characterization by Fragmentation ......................................... 449 Other Methods (Microscopy, X-ray, Electrochemistry, etc.) ...................... 451

CHAPTER 13 Measurement of Non-Humic Molecules
13.1 Introduction ........................................................................................... 453 13.1.1 Non-Humic Molecules..................................................................... 453 13.1.2 Soil Carbohydrates ......................................................................... 453 13.1.3 Soil Lipids ....................................................................................... 456 13.1.4 Pesticides and Pollutants................................................................ 457 13.2 Classical Techniques ............................................................................ 458 13.2.1 Acid Hydrolysis of Polysaccharides ................................................ 458 13.2.2 Purification of Acid Hydrolysates .................................................... 462 13.2.3 Colorimetric Titration of Sugars ...................................................... 464 13.2.4 Titration of Sugars by Gas Chromatography................................... 467 13.2.5 Quantification of Total Lipids........................................................... 472 13.2.6 Quantification of the Water-Soluble Organics ................................. 474 13.3 Complementary Techniques .................................................................475 13.3.1 Carbohydrates by Gas Chromatography..........................................475 13.3.2 Carbohydrates by Liquid Chromatography ......................................475 13.3.3 Fractionation and Study of the Soil Lipid Fraction ...........................478 13.3.4 Measurement of Pesticide Residues and Pollutants .......................483 References......................................................................................................492 Soil Carbohydrates..................................................................................... 492 Soil Lipids .................................................................................................. 494 Aqueous Extract ........................................................................................ 495 Pesticides and Pollutants........................................................................... 495

CHAPTER 14 Organic Forms of Nitrogen, Mineralizable Nitrogen (and Carbon)
14.1 Introduction ............................................................................................497 14.1.1 The Nitrogen Cycle..........................................................................497 14.1.2 Types of Methods ............................................................................499 14.2 Classical Methods..................................................................................500 14.2.1 Forms of Organic Nitrogen Released by Acid Hydrolysis ................500 14.2.2 Organic Forms of Nitrogen: Simplified Method ................................509 14.2.3 Urea Titration ...................................................................................511 14.2.4 Potentially Available Nitrogen: Biological Methods ..........................513 14.2.5 Potentially Mineralizable Nitrogen: Chemical Methods....................521 14.2.6 Kinetics of Mineralization.................................................................526 14.3 Complementary Methods ......................................................................531 14.3.1 Alternative Procedures for Acid Hydrolysis......................................531 14.3.2 Determination of Amino Acids .........................................................532 14.3.3 Determination of Amino Sugars .......................................................535 14.3.4 Proteins and Glycoproteins (glomalin).............................................538 14.3.5 Potentially Mineralizable Nitrogen by EUF ......................................538

Contents

XIII

References ......................................................................................................540 Organic Nitrogen Forms: General Articles ..................................................540 Nitrogen Forms by Acid Hydrolysis and Distillation ....................................541 Improvement of Acid Hydrolysis .................................................................541 Determination of Amino Acids ....................................................................541 Determination of Amino Sugars..................................................................542 Glomalin .....................................................................................................542 Urea Titration..............................................................................................543 Potentially Mineralizable Nitrogen: General Papers ...................................543 Potentially Mineralizable Nitrogen: Biological Methods ..............................544 Potentially Mineralizable Nitrogen: Chemical Methods...............................545 Potentially Mineralizable Nitrogen by EUF .................................................545 Mineralization Kinetics ...............................................................................546

PART 3 - INORGANIC ANALYSIS – Exchangeable and Total Elements
CHAPTER 15 pH Measurement
15.1 Introduction ........................................................................................... 551 15.1.1 Soil pH ............................................................................................ 551 15.1.2 Difficulties ....................................................................................... 553 15.1.3 Theoretical Aspects ........................................................................ 554 15.2 Classical Measurements....................................................................... 556 15.2.1 Methods .......................................................................................... 556 15.2.2 Colorimetric Method........................................................................ 557 15.2.3 Electrometric Method ...................................................................... 560 15.2.4 Electrometric Checking and Calibration .......................................... 564 15.2.5 Measurement on Aqueous Soil Suspensions ................................. 565 15.2.6 Determination of the pH-K and pH-Ca ............................................ 567 15.2.7 Measurement on Saturated Pastes ................................................ 567 15.2.8 Measurement on the Saturation Extract.......................................... 568 15.2.9 Measurement of the pH-NaF .......................................................... 569 15.3 In Situ Measurements ........................................................................... 570 15.3.1 Equipment....................................................................................... 570 15.3.2 Installation in the Field .................................................................... 570 15.3.3 Measurement on Soil Monoliths...................................................... 572 References ..................................................................................................... 574 Bibliography .................................................................................................. 575 Appendix ........................................................................................................ 576 Appendix 1: Table of Electrode Potentials ................................................. 576 Appendix 2: Constants of Dissociation of Certain Equilibriums.................. 577 Appendix 3: Buffer Solutions...................................................................... 577 Appendix 4: Coloured Indicators................................................................ 579

CHAPTER 16 Redox Potential
16.1 Definitions and Principle ...................................................................... 581 16.2 Equipment and Reagents ..................................................................... 583 16.2.1 Electrodes....................................................................................... 583 16.2.2 Salt Bridge for Connection .............................................................. 584 16.2.3 System of Measurement ................................................................. 584 16.2.4 Calibration Solutions ....................................................................... 585

XIV

Contents

16.3 Procedure............................................................................................... 585 16.3.1 Pretreatment of the Electrode ......................................................... 585 16.3.2 Measurement on Soil Sample......................................................... 586 16.3.3 Measurement on Soil Monolith ....................................................... 586 16.3.4 In Situ Measurements..................................................................... 587 16.3.5 Measurement of Oxygen Diffusion Rate ......................................... 588 16.3.6 Colorimetric Test of Eh ................................................................... 589 References ..................................................................................................... 589 Bibliography .................................................................................................. 590

CHAPTER 17 Carbonates
17.1 Introduction ........................................................................................... 593 17.2 Measurement of Total Carbonates....................................................... 595 17.2.1 Introduction ..................................................................................... 595 17.2.2 Volumetric Measurement by Calcimetry ..........................................596 17.2.3 Acidimetry........................................................................................599 17.3 Titration of Active Carbonate ................................................................601 17.3.1 Principle...........................................................................................601 17.3.2 Implementation ................................................................................601 17.3.3 Index of Chlorosis Potential.............................................................603 References......................................................................................................604

CHAPTER 18 Soluble Salts
18.1 Introduction ............................................................................................605 18.2 Extraction ...............................................................................................606 18.2.1 Soil/solution Ratio............................................................................606 18.2.2 Extraction of Saturated Paste..........................................................607 18.2.3 Diluted Extracts ...............................................................................608 18.2.4 In Situ Sampling of the Soil Water ...................................................609 18.2.5 Extracts with Hot Water ...................................................................610 18.3 Measurement and Titration ...................................................................610 18.3.1 Electrical Conductivity of Extracts....................................................610 18.3.2 In Situ Conductivity..........................................................................613 18.3.3 Total Dissolved Solid Material .........................................................614 18.3.4 Soluble Cations ...............................................................................615 18.3.5 Extractable Carbonate and Bicarbonate (Alkalinity) ........................616 18.3.6 Extractable Chloride ........................................................................618 18.3.7 Extractable Sulphate, Nitrate and Phosphate ..................................620 18.3.7 Extractable Boron ............................................................................620 18.3.8 Titration of Extractable Anions by Ionic Chromatography................622 18.3.9 Expression of the Results................................................................625 References......................................................................................................626

CHAPTER 19 Exchange Complex
19.1 Introduction ............................................................................................629 19.2 Origin of Charges...................................................................................630 19.2.1 Ionic Exchange ................................................................................630

Contents

XV

19.2.2 Exchange Complex .........................................................................631 19.2.3 Theory .............................................................................................633 References......................................................................................................636 Chronobibliography.......................................................................................637

CHAPTER 20 Isoelectric and Zero Charge Points
20.1 Introduction ............................................................................................645 20.1.1 Charges of Colloids .........................................................................645 20.1.2 Definitions........................................................................................647 20.1.3 Conditions for the Measurement of Charge.....................................649 20.2 Main Methods .........................................................................................651 20.2.1 Measurement of pH0 (PZSE), Long Equilibrium Time .....................651 20.2.2 Point of Zero Salt Effect (PZSE), Short Equilibrium Time ................652 References......................................................................................................655

CHAPTER 21 Permanent and Variable Charges
21.1 Introduction ........................................................................................... 657 21.2 Main Methods......................................................................................... 661 21.2.1 Measurement of Variable Charges ................................................. 661 21.2.2 Determination of Permanent Charges............................................. 662 References ..................................................................................................... 664 Bibliography .................................................................................................. 665

CHAPTER 22 Exchangeable Cations
22.1 Introduction ........................................................................................... 667 22.1.1 Exchangeable Cations of Soil ......................................................... 667 22.1.2 Extracting Reagents........................................................................ 668 22.1.3 Equipment....................................................................................... 669 22.2 Ammonium Acetate Method at pH 7 .................................................... 671 22.2.1 Principle .......................................................................................... 671 22.2.2 Procedure ....................................................................................... 671 22.3 Automated Continuous Extraction ...................................................... 674 References ..................................................................................................... 674 Bibliography .................................................................................................. 676

CHAPTER 23 Exchangeable Acidity
23.1 Introduction ........................................................................................... 677 23.1.1 Origin of Acidity............................................................................... 677 23.1.2 Aims of the Analysis........................................................................ 678 23.2 Method.................................................................................................... 680 23.2.1 Principle .......................................................................................... 680 23.2.2 Reagents ........................................................................................ 680 23.2.3 Procedure ....................................................................................... 681 23.3 Other Methods ....................................................................................... 683 References ..................................................................................................... 684 Chronobibliography ...................................................................................... 685

XVI

Contents

CHAPTER 24 Lime Requirement
24.1 Introduction ........................................................................................... 687 24.1.1 Correction of Soil Acidity................................................................. 687 24.1.2 Calculation of Correction................................................................. 688 24.2 SMP Buffer Method ............................................................................... 690 24.2.1 Principle .......................................................................................... 690 24.2.2 Reagents ........................................................................................ 691 24.2.3 Procedure ....................................................................................... 691 24.2.4 Remarks ......................................................................................... 692 References ..................................................................................................... 693 Chronobibliography ...................................................................................... 693

CHAPTER 25 Exchange Selectivity, Cation Exchange Isotherm
25.1 Introduction ........................................................................................... 697 25.2 Determination of the Exchange Isotherm............................................ 702 25.2.1 Principle .......................................................................................... 702 25.2.2 Reagents ........................................................................................ 702 25.2.3 Procedure........................................................................................703 25.2.4 Remarks ..........................................................................................704 References......................................................................................................705 Chronobibliography.......................................................................................706

CHAPTER 26 Cation Exchange Capacity
26.1 Introduction ............................................................................................709 26.1.1 Theoretical Aspects .........................................................................709 26.1.2 Variables that Influence the Determination of CEC..........................711 26.2 Determination of Effective CEC by Summation (ECEC) .....................718 26.2.1 Principle...........................................................................................718 26.2.2 Alternative Methods.........................................................................718 26.3 CEC Measurement at Soil pH in Not-Buffered Medium .....................719 26.3.1 Principle...........................................................................................719 26.3.2 Methods Using Not-Buffered Metallic Salts .....................................719 26.3.3 Procedure Using Not-Buffered Organo Metallic Cations .................722 26.3.4 Not-Buffered Methods Using Organic Cations ................................ 728 26.4 CEC Measurement in Buffered Medium ...............................................730 26.4.1 Buffered Methods — General Information .......................................730 26.4.2 Ammonium Acetate Method at pH 7.0.............................................732 26.4.3 Buffered Methods at pH 8.0–8.6......................................................738 26.4.4 Buffered Methods at Different pH ....................................................743 References......................................................................................................745 Bibliography ...................................................................................................750 CEC General Theory ..................................................................................750 Barium Method at soil pH ...........................................................................751 Buffered Method at pH 7.0 .........................................................................751 Cobaltihexamine CEC ................................................................................752 Silver-Thiourea ...........................................................................................753 CEC with Organic Cations (Coloured Reagents) ....................................... 753 Buffered Methods at pH 8.0–8.6.................................................................753 Barium Chloride-Triethanolamine at pH 8.1 ............................................... 753

Contents

XVII

CHAPTER 27 Anion Exchange Capacity
27.1 Theory .....................................................................................................755 27.2 Measurement ..........................................................................................758 27.2.1 Principle...........................................................................................758 27.2.2 Method ............................................................................................760 27.3 Simultaneous Measurement of AEC, EC, CEC and net CEC ..............760 27.3.1 Aim ..................................................................................................760 27.3.2 Description ......................................................................................761 References......................................................................................................763

CHAPTER 28 Inorganic Forms of Nitrogen
28.1 Introduction ............................................................................................767 28.1.1 Ammonium, Nitrate and Nitrite ........................................................767 28.1.3 Sampling Problems .........................................................................768 28.1.4 Analytical Problems .........................................................................768 28.2 Usual Methods .......................................................................................769 28.2.1 Extraction of Exchangeable Forms..................................................769 28.2.2 Separation by Micro-Diffusion......................................................... 770 28.2.3 Colorimetric Titration of Ammonium................................................ 773 28.2.4 Colorimetric Titration of Nitrites....................................................... 775 28.2.5 Colorimetric Titration of Nitrates ..................................................... 778 28.2.6 Extracted Organic Nitrogen............................................................. 779 28.3 Other Methods ....................................................................................... 780 28.3.1 Nitrate and Nitrite by Photometric UV Absorption ........................... 780 28.3.2 Ammonium Titration Using a Selective Electrode ........................... 782 28.3.3 Measurement of Nitrates with an Ion-Selective Electrode............... 785 28.3.4 In situ Measurement ....................................................................... 788 28.3.5 Non-Exchangeable Ammonium ...................................................... 790 References ..................................................................................................... 791 Bibliography .................................................................................................. 792

CHAPTER 29 Phosphorus
29.1 Introduction ........................................................................................... 793 29.2 Total Soil Phosphorus .......................................................................... 794 29.2.1 Introduction ..................................................................................... 794 29.2.2 Wet Mineralization for Total Analyses............................................. 795 29.2.3 Dry Mineralization ........................................................................... 798 29.3 Fractionation of Different Forms of Phosphorus................................ 799 29.3.1 Introduction ..................................................................................... 799 29.3.2 Sequential Methods ........................................................................ 800 29.3.3 Selective Extractions – Availability Indices ..................................... 804 29.3.4 Isotopic Dilution Methods................................................................ 813 29.3.5 Determination of Organic Phosphorus ............................................ 814 29.4 Retention of Phosphorus...................................................................... 818 29.4.1 Introduction ..................................................................................... 818 29.4.2 Determination of P Retention.......................................................... 819

XVIII

Contents

29.5 Titration of P in the Extracts................................................................. 821 29.5.1 Introduction ..................................................................................... 821 29.5.2 Titration of Ortho-phosphoric P by Spectrocolorimetry ................... 823 29.5.3 P Titration by Atomic Spectrometry . ............................................... 828 31 29.5.4 Titration of Different Forms of P by P NMR .................................. 828 29.5.5 Separation of P Compounds by Liquid Chromatography ................ 829 29.6 Direct Speciation of P in situ, or on Extracted Particles .................... 830 References ..................................................................................................... 830 Chronobibliography ...................................................................................... 833

CHAPTER 30 Sulphur
30.1 Introduction ........................................................................................... 835 30.1.1 Sulphur Compounds ....................................................................... 835 30.1.2 Mineralogical Studies...................................................................... 838 30.2 Total Sulphur and Sulphur Compounds .............................................. 839 30.2.1 Characteristics of Fluviomarine Soils .............................................. 839 30.2.2 Soil Sampling and Sample Preparation .......................................... 840 30.2.3 Testing for Soluble Sulphur Forms ................................................. 841 30.2.4 Titration of Total Sulphur................................................................. 842 30.2.5 Total S Solubilisation by Alkaline Oxidizing Fusion......................... 843 30.2.6 Total Solubilisation by Sodium Hypobromite in Alkaline Medium.... 844 30.2.7 S titration with Methylen Blue Colorimetry ...................................... 845 30.2.8 Sulphate Titration by Colorimetry with Methyl Thymol Blue............. 850 30.2.9 Total Sulphur by Automated Dry CHN(OS) Ultimate Analysis ......... 853 2– 30.2.10 Titration of Total SO4 -S by Ionic Chromatography ...................... 855 30.2.11 Total S Titration by Plasma Emission Spectrometry...................... 857 30.2.12 Titration by X-ray Fluorescence..................................................... 857 30.2.13 Titration by Atomic Absorption Spectrometry ................................ 857 30.2.14 Analytical Fractionation of Sulphur Compounds ............................ 858 30.2.15 Titration of Organic S bound to C .................................................. 859 30.2.16 Titration of Organic S not bound to C ............................................ 861 30.2.17 Extraction and Titration of Soluble Sulphides ................................ 863 30.2.18 Titration of Sulphur in Pyrites ........................................................ 865 30.2.19 Titration of Elementary Sulphur ..................................................... 867 30.2.20 Titration of Water Soluble Sulphates ............................................. 869 30.2.21 Titration of Na3-EDTA Extractable Sulphates ................................ 871 30.2.22 Titration of Jarosite ........................................................................ 873 30.2.23 Sequential Analysis of S Forms..................................................... 876 30.3 Sulphur of Gypseous Soils ................................................................... 878 30.3.1 Gypseous Soils ............................................................................... 878 30.3.2 Preliminary Tests............................................................................. 879 30.3.3 Extraction and Titration from Multiple Extracts ................................ 881 30.3.4 Gypsum Determination by Acetone Precipitation ............................ 882 30.4 Sulphur and Gypsum Requirement of Soil .......................................... 883 30.4.1 Introduction...................................................................................... 883 30.4.2 Plant Sulphur Requirement ............................................................. 884 30.4.3 Gypsum Requirement...................................................................... 886 References...................................................................................................... 888 Chronobibliography....................................................................................... 890

Contents

XIX

CHAPTER 31 Analysis of Extractable and Total Elements
31.1 Elements of Soils ...................................................................................895 31.1.1 Major Elements ...............................................................................895 31.1.2 Trace Elements and Pollutants........................................................897 31.1.3 Biogenic and Toxic Elements ..........................................................899 31.1.4 Analysis of Total Elements ..............................................................900 31.1.5 Extractable Elements.......................................................................901 31.2 Methods using Solubilization................................................................901 31.2.1 Total Solubilization Methods............................................................901 31.2.2 Mean Reagents for Complete Dissolutions .....................................903 31.2.3 Acid Attack in Open Vessel .............................................................906 31.2.4 Acid Attack in Closed Vessel...........................................................911 31.2.5 Microwave Mineralization ................................................................913 31.2.6 Alkaline Fusion ................................................................................915 31.2.7 Selective Extractions .......................................................................920 31.2.8 Measurement Methods....................................................................925 31.2.9 Spectrocolorimetric Analysis ............................................................927 31.2.10 Analysis by Flame Atomic Emission Spectrometry........................931 31.2.11 Analysis by Flame Atomic Absorption Spectrometry .....................932 31.2.12 Analysis of Trace Elements by Hydride and Cold Vapour AAS .....937 31.2.13 Analysis of Trace Elements by Electrothermal AAS ......................940 31.2.14 Analysis by Inductively Coupled Plasma-AES ..............................941 31.2.15 Analysis by Inductively Coupled Plasma-Mass Spectrometry .......946 31.3 Analysis on Solid Medium ....................................................................952 31.3.1 Method ............................................................................................952 31.3.2 X-ray Fluorescence Analysis ........................................................... 954 31.3.3 Neutron Activation Analysis ............................................................962 References .....................................................................................................969

INDEX ...................………………………………………………………………….975 PERIODIC TABLE OF THE ELEMENTS ....................................................... 993

Part 1

Mineralogical Analysis

1

Water Content and Loss on Ignition

1.1 Introduction
Schematically, a soil is made up of a solid, mineral and organic phase, a liquid phase and a gas phase. The physical and chemical characteristics of the solid phase result in both marked variability of water contents and a varying degree of resistance to the elimination of moisture. For all soil analytical studies, the analyst must know the exact quantity of the solid phase in order to transcribe his results in a stable and reproducible form. The liquid phase must be separate, and this operation must not modify the solid matrix significantly (structural water is related to the crystal lattice). Many definitions exist for the terms “moisture” and “dry soil”. The water that is eliminated by moderate heating, or extracted using solvents, represents only one part of total moisture, known as hygroscopic water, which is composed of (1) the water of adsorption retained on the surface of solids by physical absorption (forces of van der Waals), or by chemisorption, (2) the water of capillarity and swelling and (3) the hygrometrical water of the gas fraction of the soil (ratio of the effective pressure of the water vapour to maximum pressure). The limits between these different types of water are not strict. “Air-dried” soil, which is used as the reference for soil preparation in the laboratory, contains varying amounts of water which depend in particular on the nature of secondary minerals, but also on external forces (temperature, the relative humidity of the air). Some andisols or histosols that are air dried for a period of 6 months can still contain 60% of water in comparison with soils dried at 105°C, and this can lead to unacceptable errors if the analytical results are not compared with a more realistic

4

Mineralogical Analysis

reference for moisture.1 Saline soils can also cause problems because of the presence of hygroscopic salts. It is possible to determine remarkable water contents involving fields of force of retention that are sufficiently reproducible and representative (Table 1.1). These values can be represented in the form of capillary potential (pF), the decimal logarithm of the pressure in millibars needed to bring a sample to a given water content (Table 1.1). It should be noted that because of the forces of van der Waals, there can be differences in state, but not in form, between water likely to evaporate at 20°C and water that does not freeze at –78°C. The analyst defines remarkable points for example: – The water holding capacity, water content where the pressure component of the total potential becomes more significant than the gravitating component; this depends on the texture and the nature of the mineral and approaches field capacity which, after suitable drainage, corresponds to a null gravitating flow. – The capillary frangible point, a state of moisture where the continuous water film becomes monomolecular and breaks. – The points of temporary and permanent wilting where the pellicular water retained by the bonding strength balances with osmotic pressure; in this case, except for some halophilous plants, the majority of plants can no longer absorb the water that may still be present in the soil. – The hygroscopic water which cannot be easily eliminated in the natural environment as this requires considerable energy, hygroscopic water evaporates at temperatures above 100°C and does not freeze at –78°C. – The water of constitution and hydration of the mineral molecules can only be eliminated at very high pressures or at high temperatures, with irreversible modification or destruction of the crystal lattice. These types of water are estimated using different types of measurements to study the water dynamics and the mechanisms related to the mechanical properties of soils in agronomy and agricultural engineering, for example: – usable reserves (UR), easily usable reserves (EUR), or reserves that are easily available in soil–water–plant relations. – thresholds of plasticity, adhesiveness, liquidity (limits of Atterberg, etc.).

1

It should be noted that for these types of soil, errors are still amplified by the ponderal expression (because of an apparent density that is able to reach 0.3) this is likely to make the analytical results unsuitable for agronomic studies.

Water Content and Loss on Ignition

Table 1.1 Approximate correspondence moistures – pressure – diameter of the pores – types of water and critical points in soils with respect to plant requirements

5

6

Mineralogical Analysis

This brief summary gives an indication of the complexity of the concept of soil moisture and the difficulty for the analyst to find a scientifically defined basis for dry soil where the balance of the solid, liquid and gas phases is constant.

1.2 Water Content at 105°C (H2 O −)

1.2.1 Principle By convention, the term “moisture” is considered to be unequivocal. Measurement is carried out by gravimetry after drying at a maximum temperature of 105°C. This increase in temperature maintained for a controlled period of time, is sufficiently high to eliminate “free” forms of water and sufficiently low not to cause a significant loss of organic matter and unstable salts by volatilization. Repeatability and reproducibility are satisfactory in the majority of soils if procedures are rigorously respected. 1.2.2 Materials – 50 × 30 mm borosilicate glass low form weighing bottle with ground flat top cap. – Vacuum type Ø 200 mm desiccator made of borosilicate glass with removable porcelain floor, filled with anhydrous magnesium perchlorate [Mg(ClO4)2]. – Thermostatically controlled drying oven with constant speed blower for air circulation and exhausting through a vent in the top of oven – temperature uniformity ± 0.5–1°C. – Analytical balance: precision 0.1 mg, range 100 g. 1.2.3 Sample It is essential to measure water content on the same batch of samples prepared at the same time (fine earth with 2 mm particles or ground soil) for subsequent analyses. It should be noted that the moisture content of the prepared soil may change during storage (fluctuations in air moisture and temperature, oxidation of organic matter, loss or fixing of volatile substances, etc.).

Water Content and Loss on Ignition

7

This method can be considered “destructive” for certain types of soils and analyses, as the physical and chemical properties can be transformed. Samples dried at 105°C should generally not be used for other measurements. 1.2.4 Procedure – Dry tared weighing bottles for 2 h at 105°C, let them cool in the desiccator and weigh the tare with the lid placed underneath: m0 – Place about 5 g of air-dried soil (fine earth sieved through a 2 mm mesh) in the tare box and note the new weight: m1 – Place the weighing bottles with their flat caps placed underneath in a ventilated drying oven for 4 h at 105°C (the air exit must be open and the drying oven should not be overloaded) – Cool in the desiccator and weigh (all the lids of the series contained in the desiccator should be closed to avoid moisture input): m2 – Again place the opened weighing bottles in the drying oven for 1 h at 105°C and weigh under the same conditions; the weight should be constant; if not, continue drying the weighing bottles until their weight is constant

% water content at 105°C = 100 ×

m1 − m2 . m1 − m0

1.2.5 Remarks The results can also be expressed in pedological terms of water holding capacity (HC) by the soil: HC = 100 × m1 − m2 . m 2 − m0 The point of measurement at 105°C with constant mass is empirical (Fig. 1.1). A temperature of 130°C makes it possible to release almost all “interstitial water”, but this occurs to the detriment of the stability of organic matter. The speed of drying should be a function of the temperature, the surface of diffusion, the division of the solid, ventilation, pressure (vacuum), etc. Respecting the procedure is thus essential: – For andisols and histosols, the initial weighing should be systematically carried out after 6 h. – For saline soils with large quantities of dissolved salts, the sample can be dried directly, soluble salts then being integrated into the “dry soil” or eliminated beforehand by treatment with water.

8

Mineralogical Analysis

Fig. 1.1 - Theoretical diagrammatic curve showing water moved at a given temperature as a function of time (180°C = end of H2O losses in allophanes)

1.3 Loss on Ignition at 1,000°C (H2O +)
1.3.1 Introduction As we have just seen, the reference temperature (105°C) selected for the determination of the moisture content of a “dry soil” represents only _ a totally hypothetical state of the water that is normally referred to as H2O . When a sample undergoes controlled heating and the uninterrupted ponderal variations are measured, curves of “dehydration” are obtained whose inflections characterize losses in mass at certain critical temperatures (TGA).1 If one observes the temperature curve compared to a thermically inert substance (Fig. 1.2), it is possible to determine changes in energy between the sample studied and the reference substance, this results in a change in the temperature which can be measured (DTA–DSC).2 – If the temperature decreases compared to the reference, an endothermic _ peak appears that characterizes loss of H2O (dehydration), of OH (dehydroxylation), sublimation, or evaporation, or decomposition of certain substances, etc. – If the temperature increases compared to the reference, an exothermic peak appears that characterizes transformations of crystalline structures, oxidations (Fe2+ → Fe3+), etc.

2

TGA thermogravimetric analysis; DTA differential thermal analysis; DSA differential scanning calorimetry (cf. Chap. 7).

Water Content and Loss on Ignition

9

Fig. 1.2 - Schematized example of thermal analysis curves TGA (solid line) and DTA (dashed line)

The simultaneous analysis of the gases or vapours that are emitted and X-ray diffraction (cf. Chap. 4) of the modifications in structure make it possible to validate the inflections of the curves or the different endo- and exothermic peaks. As can be seen in the highly simplified Table 1.2, the most commonly observed clays are completely dehydroxyled at 1,000°C, oxides at 400°C or 500°C, carbonates, halogens, sulphates, sulphides are broken down or dehydrated between 300°C and 1,000°C, and free or bound organic matter between 300°C and 500°C. The temperature of 1,000°C can thus be retained as a stable reference temperature for loss on ignition, the thermal spectra then being practically flat up to the peaks of fusion which generally only appear at temperatures higher than 1,500°C or even 2,500°C.

1.3.2 Principle The sample should be gradually heated in oxidizing medium to 1,000°C and maintained at this temperature for 4 h.

10

Mineralogical Analysis

Table 1.2 Dehydration and dehydroxylation of some clays, oxides and salts as a function of temperature in °C

type

name

dehydrationa

dehydroxylationb

clays 1:1 clays 2:1

clays 2:1 clays 2:1 clays 2:1:1 fibrous clays

iron oxides

Kaolinite–halloysite smectites – montmorillonite Illite – micas vermiculite chlorite Sepiolite– palygorskite allophane Hematite α Fe2O3

350 370

1,000 1,000

350–370 700 600 300 200 (flat spectrum) 100 375 100 – – – – – – –

1,000 1,000 800 800–900 900–1,000 1,000 370 650 350 950–1,000 710 800–940 800 (fusion) 300 615 300–500

Al oxides Ca carbonate Mg carbonate Ca–Mg carbonate halogenous compounds sulphate sulphide organic compounds
a

goethite α FeO–OH magnetite Fe2O3 gibbsite γ-Al(OH)3 Calcite–aragonite CaCO3 magnesite MgCO3 dolomie CaMg(CO3)2 sodium chloride NaCl gypsum CaSO4, 2H2O pyrite FeS2 free or linked organic matter

b

Dehydration: loss of water adsorbed on outer or inner surfaces, with or without reversible change in the lattice depending on the types of clay, water organized in monomolecular film on surface oxygen atoms or around exchangeable cations. dehydroxylation (+ decarbonatation and desulphurization reactions), loss of water linked to lattice (OH−), irreversible reaction or destruction of the structure, water present in the cavities, O forming the base of the tetrahedrons.

Water Content and Loss on Ignition

11

Loss on ignition is determined by gravimetry. It includes combined water linked to the crystal lattice plus a little residual non-structural adsorbed water, organic matter, possibly volatile soluble salts (F−, S2−) and carbonates (CO32−, CO2). The use of an oxidizing atmosphere is essential to ensure combustion of the organic matter and in particular oxidation of reduced forms of iron, this being accompanied by an increase in mass of the soils with minerals rich in Fe2+. A complete analysis generally includes successive measurements of H2O− and H2O+ on the same sample. 1.3.3 Equipment – Platinum or Inconel (Ni–Cr–Fe) crucible with cover, diameter 46 mm. – Analytical balances (id. H2O−) – Desiccator (id. H2O−) – Muffle electric furnace (range 100–1,100°C) with proportional electronic regulation allowing modulation of the impulses with oscillation of about 1°C around the point of instruction; built-in ventilation system for evacuation of smoke and vapour – Thermal protective gloves – 300 mm crucible tong 1.3.4 Procedure – Tare a crucible, heat it to 1,000°C and cool it in the desiccator with its lid on: m0 – Introduce 2–3 g of air-dried soil crushed to 0.1 mm: m1 – Dry in the drying oven at 105°C for 4 h – Cool in the desiccator and weigh: m2 – Adjust the lid of the crucible so it covers approximately 2/3 of the crucible and put it in the electric furnace – Programme a heating gradient of approximately 6°C per minute with a 20-min stage at 300°C, then a fast rise at full power up to 1,000°C with a 4-h graduation step (the door of the furnace should only be closed after complete combustion of the organic matter) – Cool the crucible in the desiccator and weigh: m3 1.3.5 Calculations m1 – m0 = m1 – m2 = weight of air-dried soil moisture at 105°C

12

Mineralogical Analysis

m2 – m0 = m2 – m3 =

weight of soil dried at 105°C loss on ignition
m1 − m2 m1 − m0 m2 − m3 m2 − m0

H 2O – % = 100 × H 2O % = 100 ×
+

related to air-dried soil related to soil dried at 105°C

1.3.6 Remarks Knowing the moisture of the air-dried soil, it is possible to calculate the weight of air-dried soil required to work with a standard weight soil dried at 105°C, thus simplifying calculations during analyses of the samples. To obtain the equivalent of 1 g of soil dried at 105°C, it is necessary to weigh:

100 with wc = % water content of air dried soil. 100 − wc
Platinum crucibles are very expensive and are somewhat volatile at 1,000°C, which means they have to be tared before each operation, particularly when operating in reducing conditions. Combustion of organic matter with insufficient oxygen can lead to the formation of carbide of Pt, sulphides combine with Pt, chlorine attacks Pt, etc.

Bibliography
Campbell GS, Anderson RY (1998) Evaluation of simple transmission line oscillators for soil moisture measurement. Comput. and Electron. Agric., 20, 31–44 Chin Huat Lim, Jackson ML (1982) Dissolution for total elemental analysis. In Methods of Soil Analysis, Part 2, Page A.L., Miller R.H., Kenny D.R. ed. Am. Soc. Agronomy, pp. 1–11 Dixon JB (1977) Minerals in soil environments. Soil Sci. Soc. Am. Dubois J, Paindavoine JM (1982) Humidité dans les solides, liquides et gaz. Techniques de l’ingénieur., (P 3760) Gardner WH (1986) Water content. In Methods of Soil Analysis, Part 1, Klute ed. Am. Soc. Agronomy, Soil Sci. Soc. Am., pp. 493–544 Henin S (1977) Cours de physique du sol: l'eau et le sol tome II., Editest, Paris: 1–164 Lane PNJ, Mackenzie DH, Nadler AD (2002) Note of clarification about: Field and laboratory calibration and test of TDR and capacitance techniques for indirect measurement of soil water content. Aust. J. Soil Res., 40, 555–1386

Water Content and Loss on Ignition

13

Lane PNJ, Mackenzie DH (2001) Field and laboratory calibration and test of TDR and capacitance techniques for indirect measurement of soil. Aust. J. Soil Res., 39, 1371–1386 NF ISO 11465 (X31-102) (1994) Détermination de la teneur pondérale en matière sèche et en eau. In Qualité des sols, AFNOR, 1996, 517–524 Rankin LK, Smajstrla AG (1997) Evaluation of the carbide method for soil moisture measurement in sandy soils. Soil and Crop Science Society of Florida, 56, pp. 136–139 Skierucha W (2000) Accuracy of soil moisture measurement by TDR technique. Int. Agrophys., 14, 417–426 Slaughter DC, Pelletier MG, Upadhyaya SK (2001) Sensing soil moisture using NIR spectroscopy. Appl. Eng. Agric., 17, 241–247 Walker JP, Houser PR (2002) Evaluation of the Ohm Mapper instrument for soil moisture measurement. Soil Sci. Soc. Am. J., 66, 728–734 X31-505 (1992) Méthode de détermination du volume, apparent, et du contenu en eau des mottes. In Qualité des sols, AFNOR, 1996, 373–384 Yu C, Warrick AW, Conklin MH (1999) Derived functions of time domain reflectometry for soil moisture measurement. Water Resour. Res., 35, 1789–1796

2 Particle Size Analysis

2.1 Introduction

2.1.1 Particle Size in Soil Science Determination of grain-size distribution of a sample of soil is an important analysis for various topics in pedology, agronomy, sedimentology, and other fields such as road geotechnics. Soil texture has an extremely significant influence on the physical and mechanical behaviours of the soil, and on all the properties related to water content and the movement of water, (compactness, plasticity, thrust force, slaking, holding capacity, moisture at different potentials, permeability, capillary movements, etc.). Particle size analysis of a sample of soil, sometimes called “mechanical analysis”, is a concept that has been the subject of much discussion (Hénin 1976). Soil is an organized medium including an assemblage of mineral and organic particles belonging to a continuous dimensional series. The first difficulty is to express the proportion of these different particles according to a standard classification, which is consequently somewhat artificial. One classification scale was proposed by Atterberg (1912). Today this scale is recognized at different national and international levels and includes two main fractions: fine earth (clay, silts and sands with a grain diameter <2 mm) and coarse elements (gravels, stones with a grain diameter >2 mm). The particle size series (Fig. 2.1) for fine earth is generally expressed after analysis in three size fractions (clay fraction less than 0.002 mm, silt fraction from 0.002 to 0.02 mm, and sand fraction from 0.02 to 2 mm). In some countries, or for the purpose of a particular type of pedological interpretation, a more detailed scale of classes is sometimes used, for example five fractions: fine clays, silts, coarse silts or very fine sands, fine sands, and coarse sands (Fig. 2.1).

16

Mineralogical Analysis

Fig. 2.1. Ranges of particle size used for soils (NC number of classes; FSi fine silts, CSi coarse silts; FS, VFS, CS fine, very fine and coarse sands, respectively; FC fine clays; FG, CG fine gravels and coarse gravels), from top to bottom: (CSSC) Canadian Soil Survey Committee (1978): 10 particle size ranges < 2 mm; France (before 1987): 8 ranges; USDA United States Department of Agriculture (1975): 7 ranges; AFNOR Association Française de Normalisation (1987): 5 ranges; ISSS = International Soil Science Society (1966): 4 ranges; ASTM = American Society for Testing Materials (1985): 3 ranges

However, it should be noted that the terminology used does not provide much information about the real nature of the classes; thus clay defined as having a diameter equal to or less than 0.002 mm does not contain only clay corresponding to this mineralogical definition but can also contain sesquioxides, very fine silts, organic matter, carbonates, or compounds without colloidal properties. In the same way, sands, which generally result from fragmentation of the parent rock, can also include pseudo-sands, small ferruginous concretions, small limestone or cemented nodules that are resistant to dispersion treatments. The presence of these pseudo-sands can render the conclusions of particle size analysis illusory. Another difficulty appears with the fractionation of elementary particles by dissociating them from their original assembly. Here too analytical standards exist, but it should be recognized that in certain cases the rupture of all the forces of cohesion is not complete (the case in hardened cemented soils), or on the contrary the forces are too energetic. Lastly, particle size analysis accounts for the size but not for the shape of the particles, or their nature. If necessary, these are the subject of

Particle Size Analysis

17

specific morphoscopic and mineralogical analyses. The result of particle size analysis is expressed in classes of which the relative proportions can be summed up in the form of a triangular diagram enabling the texture of a sample, a horizon, or a soil to be defined. Depending on the school, there are several different types of triangles that represent textures: GEPPA (Groupe d’Etude des Problèmes de Pédologie Appliquée, AFES, Grignon, France) includes 17 textural classes; the USDA’s (United States Department of Agriculture) includes 12 classes (Gras 1988); others are simplified to a greater or lesser extent depending on the pedological or agronomic purpose of the study. Starting from these results, different interpretations are usually made in terms of pedogenesis (comparison of the vertical sand percents to check the homogeneity of a given material in a given soil profile, calculation of different indices of leaching, clay transport, etc.); others are more practical (definition of the relation of texture to hydric characteristics for the initial calculation of the amounts and frequencies of irrigation, or for the choice of machinery for cultivation. 2.1.2 Principle Particle size analysis is a laboratory process, which initially causes dissociation of the material into elementary particles; this implies the destruction of the aggregates by eliminating the action of cements. But this action should not be too violent to avoid the creation of particles that would not naturally exist; the procedure of dispersion must thus be sufficiently effective to break down the aggregates into individual components, but not strong enough to create neo-particles. Measurements (Table 2.1, Fig. 2.2) then will link the size of the particles to physical characteristics of the suspension of soil after dispersion (cf. Sect. 2.1.3). These measurements may be distorted by the presence of some compounds in the soil: organic matter, soluble salts, sesquioxides, carbonates, or gypsum. The latter compound can be particularly awkward because it can result in two opposing actions (Vieillefon 1979): flocculation due to soluble calcium ions (relative reduction in clay content), and low density of gypsum compared to other minerals (increase in clay content). Particle size analysis thus generally starts with a pre-treatment of the sample that varies with the type of soil; the characteristics of different soils are given in Table 2.3.

18

Mineralogical Analysis

– – – – –

– – – – –

– – – – – – – – – – – – – – – – –

Fig. 2.2. Particle size ranges of some automated particle-measurement instruments

2.1.3 Law of Sedimentation After possible pretreatment (cf. Sect. 2.2.1), the sample is suspended in aqueous medium in the presence of a dispersant (cf. Sect. 2.2.2). During sedimentation, the particles are then subjected to two essential forces: a force of gravity that attracts them to the bottom, and a force of viscous resistance of the medium in the opposite direction to their displacement. By comparing the particles to spheres of radius r, the force of gravity Fg (dynes) is expressed by:
Fg = 4 3

p r 3 ρs − ρ f g

(

)

r = equivalent radius of the spherical particle in cm; g = gravity constant, 981 cm s–2; ρs = density of the particles in g cm–3 (between 2.4 and 2.8 for soils); ρf = density of the liquid of dispersion in g cm–3; The force of resistance of the medium Fr (dynes) is expressed by:

Particle Size Analysis

19

Fr = 6 p r η V ,

V = falling speed in cm s–1; η = viscosity of the medium in Poises (g cm–1 s–1), at temperature θ °C (Table 2.2). When the particles reach equilibrium, the forces Fr and Fr are equal, from which their drop speed can be estimated according to the law originally established by Stokes (1851):
V= 2 (ρs −ρf ) g r 9η
2

.

(2.1)

For calculations, the average density of the solid particles in dispersions of soils is often selected with ρ S = 2.65 or 2.60 g cm–3. Empirical relationships have been established for the calculation of ρF and η in aqueous solutions of hexametaphosphate generally used for particle-size distribution of soils (Gee and Bauder 1986):

ρt = ρ 0 (1 + 0.630 CHMP), η = η0 (1 + 4.25 CHMP),

(2.2) (2.2’)

CHMP = hexametaphosphate concentration in g cm–3. The constant of Stokes for the medium can thus be defined by: C = 2 (ρs – ρf ) g/9 η.

ρ0 = density of water (g cm–3 at the working temperature (Table 2.2); η0 = viscosity of water (poise) at the working temperature (Table 2.2);

Equation (2.1) shows that the falling speed is proportional to the square of the particle radius and remains constant throughout sedimentation if certain conditions are strictly respected (cf. Sect. 2.1.4). The speed can also be defined by V = h/t where T is the time (s) spent by the particle of radius r(cm) to fall a height H(cm). Either the depth of its sedimentation over a given period, or the time needed for sedimentation to a given depth is determined by:
t = 2 ρs − ρ f g r

(

9 h η

)

2

= h C

−1

r

−2

.

(2.3)

20

Table 2.1. Systems of particulate characterization for particle size distribution of soils

individualization of particles

suspension (dispersion)

– destruction of organic matter (H2O2, Na hypochlorite), hypobromite... – destruction of cements (Al, Fe, Si): – acid or basic media – reducer or complexing media – in water 2+ desaturation (elimination cations ) chemical acid medium (some andosols) choice of pH basic medium: NaOH, NH4OH, pyrophosphate, hexametaphosphate – preliminary treatments various surfactants ultrasounds physical mechanical agitation (disintegration: 40 reversals/min) limiting concentrations – choice of concentration wall-attachment effects phase recovery principle advantages drawbacks 1990 firms

separation – techniques used – measurements

size range

measurement of yes

dry

1. sieving

particles

wet

2 mm – 0.050 (5 µm)

fragility of sieves, mesh defects, mesh obtrusions, etc. when dry, fine powders stick to coarse ones

Saulas Tamisor, etc

Mineralogical Analysis

2. surface measurement (for memorandum)

≤ 2 µm no

measurement by separation on sieve with vibration with or without ultrasonic waves simple weighing of the fractions (discontinuous) measurement of weight of a molecular film (nitrogen, EGME...) retained at internal and the surface of the external surface particles (preliminary separation of phases)

difficult to measure

micromeritics

dry yes

density

very slow used for sediment studies

3. elutriation

soils = 2.65

Particle Size Analysis

wet

1– 150 µm

separation measurement inverse of sedimentation: breakdown of gravity force by a gaseous or liquid flow Per Ascensum (discontinuous)

radioactivity (for memorandum) for labelled elements no no

by fluctuation of conductivity (or photometrics)

hole calibration, obturation

Coulter

dry or wet samples no

4. counting

0.5 – 500 µm

by microscopic measurement: optical, MEB, MET… and image analysis

calibrated hole, distribution of given particles by elimination of proportional changes density effect in displaced electrolyte or proportional extinction shape, habitus counting on image analyzer or manual counting

equipment is expensive

Quantimat 720 MicroVideomat (Zeiss) Integramat (Leitz)

sedimentation balances no

1– 150 µm

gravimetric determination by cumulating vs. time

very slow analysis

Mettler, Becker, Cahn, Sartorius, Prolabo

21

22

separation techniques used – measurement principle firms advantages drawbacks

size phase range recovery

1990

1– 150 µm yes

volumetric sampling of fine particles. sieving for silts and sands

sedimentation at fixed level vs. time. Critical concentration per unit of volume

pipette methods (Robinson, Andreasen) reference method AFNOR standardization simple gravity photosedimentation (turbidimetry) diffusion diffraction laser. no optical method: absorption, diffusion or dispersion of light (white or monochromatic light or laser) influence of shape of colour of particles on laser: high extinction intensity lighting coefficient, of small volume influence of low concentration wavelength diffraction effect for small particles

1 – 20 – 50 µm

Stanton Malvern... Cilas

5. sedimentation

only one no

X-ray soft γ attenuation neutron scattering measurement by electromagnetic radiation – X-ray or γ

0.1 – 100 µm

good detection

concentration higher than photometry

Micromeretics 5100...

method cannot

Mineralogical Analysis

be used for

density variable depth

1 – 50 µm

bouyoucos: measurement of densimeter depth

densimeter (Casagrande, ASTM...)

influence of density, temperature, viscosity, verticality of cylinders

Particle Size Analysis

hydrometer constant depth De Leenher: densimeter with chain surface evaporation

all No

continuous centrifugation

classes of

5’. centrifugation

0.02 – 500 µm

increase in gravity force by centrifugation force. Change of concentration at a given depth, photo detection

Shimadzu Capa 500 Horiba 300 Horiba 700 Horiba Beckman

discontinuous centrifugation

particles = yes

Stokes law: speed of solid fractionation: one static liquid phase – one solid mobile phase, no deformation, no reaction with liquid phase, homogeneously dispersed. Effect of temperature (viscosity, convection flows), effect of density, particle diameter. Repulsive electrostatic effect of particles

0,02 – 2 µm

– weighing of cumulated deposits or measurement of height of deposits – possible approximate separation of phases – pipette centrifuge analyser

Sharples Simcar ...

6) M Brownian Zeta potential

movement

0.003 – 2 µm

the frequency of displacements is inversely proportional to particle size

mobility of particles, effective surface charge

fine particles only

Malvern

23

24

Mineralogical Analysis

Particle size analysis by sedimentation consists in determining the content of particles below or equal to a given threshold. Known volumes of solution (pipette method) are generally used for the depth and time of sedimentation chosen as the threshold for a cut point. After drying the pipette sample, weighing and correcting the volume, the content of particles that are smaller than the selected threshold can be determined. In the example in Table 2.2, a pipette sample at a temperature of 20°C, a depth of 10 cm and 8 h 08 min of sedimentation will give the content of the clay fraction (diameter of particle < 2 µm). In the densimetry method, the relation between the size of the particles (radius r) and the time of sedimentation t can be expressed by: r = S t –1/2, (2.4) where S is the parameter of sedimentation. Taking into account (2.3), it can be expressed by: S = C –1/2 Hr1/2, (2.4’) where Hr is the depth of balance of the densimeter (hydrometer) which represents the effective measurement depth of the particles of radius r.

2.1.4 Conditions for Application of Stokes Law The formula of Stokes is theoretically only valid for particles with a diameter of less than 0.1 mm, but according to Mériaux (1954), it can be used up to 0.2 mm or even 0.05 mm. Above this value, it is advisable to apply the formula of Oseen; however, particles more than 0.1 mm in diameter can be more precisely sorted by sieving. For particle size analysis, sedimentation cylinders are used whose walls slow down the falling speed of the particles by friction. Thus, for 0.05 mm quartz spheres, the falling speed at a distance of 0.1 mm from the walls is reduced by 12%, and disturbance becomes negligible at 1 mm (0.28%). In practice, it may be advantageous to use sedimentation tubes (cylinders) with a rather large diameter, at least 5 cm. In addition, the constant of Stokes is established for minerals with an average density of 2.60 or 2.65, whereas soil materials can contain illite with a density of 2.1 – 2.7, montmorillonite with a density of 1.7 – 2.6 and so on. But the main difficulty is the fact that the particles are neither spherical, nor smooth, which obliges the analyst to introduce the concept of equivalent radius.

Particle Size Analysis

25

Table 2.2. Densities ρ and viscosities η of water and 5% hexametaphosphate solutions related to temperature θ (°C); corresponding to Stoke C constants and falling time t at 10 cm depth for clay particles ≥ 2 µm –3 (ρs = 2.60 g cm ) in hexametaphosphate solution
t (2.3) for h = 10 cm r = 0.0001 cm CHMP = 0.05 9 h 15 min 9 h 00 min 8 h 46 min 8 h 33 min 8 h 20 min 8 h 08 min 7 h 56 min 7 h 45 min 7 h 34 min 7 h 23 min 7 h 13 min

θ

ρ water

η water Poise

ρ corrected η corrected
C (2.2; (2.2’; CHMP = 0.05 CHMP = 0.05 (eau) g cm–3) g cm–3)

C (CHMP = 0.05 g cm–3)

(°C) (g cm–3) 15 16 17 18 19 20 21 22 23 24 25

(g cm–1 s–1) (g cm–3) 1.030598 1.030438 1.030264 1.030080 1.029884 1.029676 1.029459 1.029230 1.028990 1.028741 1.28482

(g cm–1 s–1) (cm–1 s–1) (cm–1 s–1) 0.01381 0.01345 0.01311 0.01277 0.01245 0.01215 0.01186 0.01158 0.01131 0.01105 0.01080 30640 31472 32290 33153 33996 34849 35712 36581 37462 38347 39245 30038 30853 31656 32502 33329 34165 35012 35864 36727 37596 38476

0.999126 0.01139 0.99897 0.01109

0.998802 0.01081 0.998623 0.01053 0.998433 0.01027 0.998232 0.01002 0.998021 0.009779 0.997799 0.009548 0.997567 0.009325 0.997325 0.009111 0.997074 0.008904

Particle size analysis is concerned with sedimentation of particles of different sizes. Some particles sediment more quickly than others and this results in variations in viscosity during the course of the experiment and also variations in the density of the fluid. Thus, in order to not diverge too much from the theoretical conditions established for mono-dispersed systems, a too significant concentration of the soil sample should be avoided (never higher than 1%). The graph of the sedimentation of a heterogeneous sample corresponds to a poly-dispersed system. This construction (Fig. 2.3) makes it possible to evaluate the percentage of particles with a diameter larger than a value “A” corresponding to a sedimentation time ‘tx’.

26

Mineralogical Analysis

A to tx

Fig. 2.3. Curve of sedimentation of a complex sample (poly-dispersed system); tx is time representing the value of the sum of particles X. The intersection of the tangent at point X of the curve with the y-coordinate gives in A the quantity of particles larger than X

2.2. Standard Methods
2.2.1 Pretreatment of the Sample Adaptation to Soil Type The sample is dried at room temperature, sieved to 2 mm and carefully partitioned with a manual riffle sampler (Pansu et al. 2001). The weight of the test sample is 10 g; this quantity can be increased to 20 g for soils which are not very rich in clays. The treatments below are defined with respect to the different cases (A – H) listed in Table 2.3.

Particle Size Analysis

27

Carbonated Soils (B) Reagents HCl 1 mol L–1

Table 2.3. Characteristic of soils subjected to analysis and preliminary treatments
organo-mineral complex characterization treatment

A: complex more or less saturated

few drops of concentrated HCl do not cause any release of CO2, pH H2O < 7

treatment with H2O2 at 1%, then 30%

B: presence of carbonates C: rich in organic matter D: rich in organomineral cements and amorphous or individualised sesquioxides E: rich in sodic salts, no gypsum

HCl causes release of CO2 perform two analyses: one without carbonate destruction, and one pH H2O > 7 with destruction OM content > 15% dried <2mm thin earth (SiO2 + Al2O3) > 10% of thin earth – ferrallitic, ferruginous tropical soils suitable treatment with H2O2 – dissolution of mineral cements with HCl and organic cements with H2O2 – treatment with Tamm reagent conductivity 1/5 > 0.3 mS cm–1; to 5 mL of aqueous extract, add 5 mL acetone – no precipitate strong reaction with H2O2; violet colour of permanganate after oxidation elimination of salts by washing and decantation–filtration

F: presence of MnO2

pretreatment with bisulphite before destruction of organic matter

G: presence of gypsum < 10%

positive reaction to acetone elimination of gypsum by washing test before agitation estimation of gypsum by heating to 60°, then to 105°C (cf. Sect. 2.2.4) results difficult to interpret. Special techniques should be used

H: presence of gypsum > 10%

28

Mineralogical Analysis

Procedure The analytical standard X 31-107 (1983) recommends carrying out this treatment after destruction of organic matter with hydrogen peroxide (H2O2). However, oxidation of the organic matter can temporarily produce oxalic acid which results in the neo-formation of calcium oxalate. It is thus preferable to eliminate the limestone at the pretreatment stage. − Put the sample (equivalent to 10 g soil dried at 105°C for example) weighed at ± 1 mg in a low form 1,000 mL Pyrex beaker. − Add 100 mL water. − Agitate and with a burette slowly add the diluted hydrochloric acid (1 mol L–1) until complete destruction of the limestone present. − The pH should not go below 3.0. When all carbon dioxide is eliminated, boil gently for 5 min. Wash by decantation to eliminate calcium and excess acid (chloride test). Remarks − Avoid adding too much hydrochloric acid which can destroy the chlorites in certain trioctaedric chlorites. − Sodium acetate at pH 5 can be used to avoid attacking the lattice of certain clays. − In the presence of limestone particles involved in the grain-size distribution of the soil, an additional measurement will be required without destroying the CaCO3 (Baize 2000). Destruction of Organic Matter Organic matter has a high aggregation capacity. It should thus be destroyed in the majority of soils (A – F in Table 2.3). Generally hydrogen peroxide at 30% (110 volumes) is used or stabilized hydrogen peroxide (Perhydrol or similar) in tropical climates. Some authors propose sodium hypochlorite or bromine in an alkaline medium (2 mol (KOH) L–1 solution). Reagents – Pure hydrogen peroxide (30% – 110 volumes) – Ammonia (20%, d = 0.92) – Dispersant, 5% sodium hexametaphosphate solution – Sodium bisulphite NaHSO3 Case of Soils with Less Than 15% of Organic C Place the test specimen in a 1,000 mL Pyrex beaker. Add 100 mL of 1%. hydrogen peroxide. Leave in contact in the cold for one hour avoiding

Particle Size Analysis

29

excess foaming by agitating, or by using an aerosol to modify the surface charges (alcohol, etc.). Heat to 60°C and add a little 30% hydrogen peroxide to start the attack again. Add H2O2 in small fractions until effervescence stops and there is discolouration of the supernatant. Bring to controlled boiling to destroy surplus H2O2 and to reduce the volume without bringing to dry. Case of Soils Very Rich in Organic Matter (Histosols, Andosols, etc.) It is important to work on soils preserved in their natural moisture to avoid irreversible changes due to drying as these soils become hydrophobic. A sample equivalent to soil dried at 105°C should be used. The attack must be very gentle at the beginning because as soon as the oxidation reaction starts, it becomes violent; there is a sudden rise in temperature and a risk of overflow of foam. When sampling wet soil, add distilled water to form slurry. Add 50 mL hydrogen peroxide diluted to 1% and leave in contact in the cold. Heat each beaker to 60°C to start the reaction. If necessary, adjust the temperature by adding an ice cube made of deionized water. Add small fractions of hydrogen peroxide until there is no more foam, then bring to the boil. The liquid supernatant should be clear. Wash by decantation and continue the analysis. Remarks In certain cases, the organic matter may be “protected” by homogeneous mixture with carbonates and hydrogen peroxide then cannot act. If preliminary destruction of the carbonates (cf. “Carbonated Soils”) is not sufficient, the organic matter should be attacked with hypobromite: 3 mL of bromine in 100 mL of frozen 2 mol (KOH) L–1 solution. Mix 50 mL of this mixture with the sample and wait 1 h; boil for 30 min, let cool, add 200 mL distilled water; leave overnight, transfer on a filter and wash (three washings are sufficient) before dispersion. Presence of MnO2 (F) The presence of manganese salts can cause the rapid destruction of hydrogen peroxide. In this case the treatments should be renewed and the colour of the supernatant liquid monitored. Free manganese dioxide is soluble in hydrogen peroxide (Jackson 1969). Since manganese dioxide causes violent breaking up of hydrogen peroxide, it should first be reduced with sodium bisulphite (Pétard, 1993). Before adding hydrogen peroxide, add 1 g of sodium bisulphite or 10 mL of an aqueous bisulfite solution at 37.5% to the sample. Add 50 mL

30

Mineralogical Analysis

deionized water and boil for 20 min to reduce the manganese dioxide in Mn2+ ion. Then initiate the attack with hydrogen peroxide as described above. Presence of Amorphous Organo-Mineral Cements (D) Reagents − Tamm (1922) buffer: 10.92 g of oxalic acid + 16.11 g of ammonium oxalate for 1,000 mL; adjust to pH 3; − Hydrochloric acid 2 mol L–1 solution: dissolve 166 mL concentrated HCl (d = 1.18) in 800 mL water, agitate, cool and complete to 1 L. Procedure with Tamm Reagent Add 800 mL of Tamm reagent to a sample weighing 20 g; agitate cold, store in the dark for 4 h, centrifuge, and then filter. Procedure with HCl/H2O2 Reagent Treat with 300 mL 2 mol (HCl) L–1 solution in a sand bath for 1 h at 60°C. Elutriate and wash with deionized water. The acid solution and washing water should be collected and dried at 105°C. Weighing gives the mass Mm of the soluble mineral fraction. If m represents the initial test specimen of soil, the mineral soluble fraction Fm expressed as a percentage is calculated by Fm = 100 Mm /m. This value should be taken into account for the calculation of the balance of particle size fractionation. After dissolving mineral cements, organic cements should be dissolved as described in above “Destruction of Organic Matter”. Presence of Gypsum (G and H) Procedure for Rough Estimate of Gypsum Put a test specimen of 10 g of soil sieved to 2 mm in an aluminium capsule. Place in a ventilated drying oven at 60°C for 24 h to eliminate the water of hydration. Cool in the desiccator and weigh = P1; place in a ventilated drying oven regulated at 105°C (for minimum 3 h) to eliminate the water of constitution; cool in the desiccator and weigh = P2. P1− P2 172 × ≈ 50(P1 – P2). Approximate percentage gypsum = 100 10 36

Particle Size Analysis

31

Procedure in Case of Gypsum ≤ 10% (case G) At 25°C, gypsum is water soluble at a rate of 2 g L–1. After destruction of organic matter (cf. “Destruction of Organic Matter”) and after destruction or not of limestone (cf. “Carbonated Soils”), put the sample (10 g) in a 500 mL beaker with 300 mL distilled water on a magnetic stirrer with agitation. After 1 h, allow it to settle and elutriate the clear part, again add 300 mL distilled water and repeat the operation (usually once again) until the acetone test is negative. For this test (cf. Chap. 30), use 5 mL aqueous extract, add 5 mL acetone, mix well; in the absence of gypsum, no precipitate will be formed. Procedure in Case of Gypsum > 10% (case H) Determination of the exact particle size is difficult and it is advisable to adapt a specific method, for example, that of Vieillefon (1979). Complete dissolution of the gypsum enables the elementary composition of the nongypseous part of the soil to be determined, but not of the total soil. Instrumental methods are more precise because only a short measurement time is required. Combined with desaturation using ion exchange resins, these methods make it possible to obtain satisfactory results before flocculation occurs. 2.2.2 Particle Suspension and Dispersion Introduction Ions that can flocculate clay particles are ranked in descending order: Al3+ > Ca2+ > NH4+ > K+ > Na+ > Li+ The replacement of the natural compensation cations with high flocculating capacity is thus necessary to enable dispersion of the soil, i.e. the maintenance of the elementary particles in suspension. The stability of the suspensions is only obtained thanks to interactions between the diffuse layers of the same sign as clays and with control of the forces of attraction of van der Waals. Coulomb repulsion depends on the concentration of electrolytes and the valence of the ions. If the forces of van der Waals are higher than the force of Coulomb repulsion, the potential energy of the resulting interaction leads to flocculation. The aim is to chemically modify the distribution of charges and pH. For example, inversion of the edge charges can eliminate edge lattice attractions and produce negative particles with a weak attraction potential. For example hexametaphosphate is fixed by chemisorption on

32

Mineralogical Analysis

the octahedral cations present on the side faces of crystallites resulting in good dispersion. This dispersant is most often used when analyses are not required on the fractions after measurement of particle size. The pH should preferably be fixed at a value which is far from the zone of the isoelectric point (cf. Chap. 20) at which flocculation occurs. Generally it should be basic (approximately 10). A too high pH can induce solubilization phenomena. For oxides like Fe(OH)3, which can exist in a dispersed state with a positive or negative charge, adjustment of the pH to allow dispersion is a function of the isoelectric point. For Andisols, the pH for dispersion will often be acid (around pH 3.5) and always far from the isoelectric point. These treatments do not avoid lattice associations which are observed in kaolinites with a low negative charge or in micas (illites, muscovite, etc.). In this case it is necessary to use ultrasound for an additional mechanical effect.1 The optimal amount of dispersant will depend on the effectiveness of the pre-treatments. Optimal effectiveness is obtained with an amount of sodium hexametaphosphate of between 20 and 50 times the CEC. An excess of this reagent should be avoided as it causes flocculation while being adsorbed on colloids. For this reason, and when the fractions are needed for later analysis, it is better to use ion-exchange resins in Na+ form as dispersants, e.g. Amberlite IR 120 Na (Rouiller et al.1972). These have the advantage of not causing any additional ionic charge of the medium, which is very favourable in the case of horizons with low clay content where a saline medium adds significant weight to the clay content. Among the dispersants that can be used are ammonia, soda, sodium carbonate, and sodium pyrophosphate. All these agents avoid compression of the double layer of crystallites and avoid raising the Zeta
1

Ultrasounds: (see complementary bibliography for effects on soils). Ultrasonic vibrations are generated by magnetostrictive oscillators. When a bar of a ferromagnetic material is subjected to a magnetic field, it changes length by magnetostriction. When this alternative field is applied in the axis of the bar, this causes an oscillation that is double of the applied field frequency. This vibration is transmitted to the suspended particles by the aqueous medium. The effect of cavitation with a frequency from 20 to 30 kHz makes it possible to break the forces of cohesion of the aggregates without causing significant damage to the elementary particles, as long as the application time is short. The treatment causes a rise in temperature which should be controlled. The apparatus are built so as to avoid the zones of resonance waves which are most destructive. Two types of apparatus are used, either with tanks or with probes with mechanical agitation by blades. Agitation with a bar magnet is not used except to recover magnetic particles if required.

Particle Size Analysis

33

potential, thus maintaining inter-particle forces of repulsion. It should be noted that at a rate of 17 mL per litre, ammonia does not cause an overload for weighing, and that soda dissolves the organic matter and precipitates iron, whereas pyro- and hexametaphosphate maintain it in solution. Aluminium is put in solution in the form of aluminates with soda, but not with ammonia. The responses are thus always slightly different. As the mixed hexametaphosphate–ammonia dispersant has one component with a higher density than water and one component with a lower density, the resulting density is close to that of water. Equipment and Reagents − Sedimentation cylinders. graduated 1,000 mL cylinders with ground stoppers (45/40 mm), 400 mm in length and 60 mm in diameter are generally used, but equivalent results can be obtained using transparent PVC tubes with an internal diameter of 71 mm and a length of 30 cm, with a filling mark at 1,000 mL, a square base, and stopped with a rubber stopper for agitation. − Dispersing solution. 15% sodium hexametaphosphate solution (calgon, (NaPO3)6) in deionized water. − 20% ammonia (d = 0.92) solution. − Ion exchange resin (Amberlite IR120 Na or similar). Procedure Treatment Using the Mixed Dispersant Hexametaphosphate – Ammonia After subjecting the soil sample to suitable pretreatments, quantitatively place it on an unfolded analytical filter and wash it with deionized water until dispersion begins. Pierce the filter and with jets of water from a washing bottle direct the soil into a cylinder (Fig. 2.4). Add 10 mL of 15% sodium hexametaphosphate solution and 5 mL of 20% ammonia solution. Supplement with approximately 500 mL water, close and place on the rotary agitator for 2 h (4 h for clay soils). In the case of soils of andosol–histosol type, first carry out ultrasonic treatment for 15 min). After agitation, the sample should be well dispersed and its elementary particles (sands, silts, clays) quite separate from each other. After a few minutes, check there is no flocculation. Bring the volume to 1,000 mL with deionized water and homogenize.

34

Mineralogical Analysis

Treatment Using Na Form IR 120 Resin Before treatment, the resin should be removed from particles smaller than 200 µm by sieving. Place samples that have been subjected to suitable pretreatment quantitatively on a filter and wash with deionized water until dispersion begins. Pierce the filter and place the soil on a 200-µm mesh sieve in a funnel on a sedimentation cylinder to recover coarse sands. Wash, dry, and filter the recovered sands (cf. “Washing and Measuring Fine and Coarse Sands”). Place the fine particles in the cylinder. Add 50 mL wet Na form resin, then approximately 500 mL deionized water and agitate with the rotary agitator for 4 h (5 h for soils containing < 10% gypsum). After agitation, recover the resin on a 200 µm mesh sieve placed in a funnel in a 1,000 mL sedimentation cylinder, then wash separately and recover the washing water (the resin can be regenerated for another operation). The sample can also be brought to 1,000 mL directly with deionized water for sampling of the fine particles, which can be needed for later analyses. It is alsopossible to add 10 mL of 15% sodium hexametaphos phate solution and 5 mL 20% ammonia, as in the procedure described in “Treatment Using the Mixed dispersant Hexametaphosphate – Ammonia”, the densities and viscosities are then similar.

Fig.2.4. Transfer in sedimentation cylinders for dispersion after pretreatment, filtration and washing

Particle Size Analysis

35

2.2.3 Pipette Method after Robinson-Köhn or Andreasen Principle The pipette method is based on sedimentation of the particles by gravity according to the law of Stokes (2.1). Recovery of the aliquot at a given depth and a given time makes it possible to identify a specific class of particles when all the particles bigger than the selected diameter have been eliminated. Equipment – Sand bath. – Robinson pipette on moveable frame with toothed rack (Fig. 2.5), aspiration by micropump with flow regulated at 60 mL min–1. The pipette should have undergone preliminary treatment to make it non-wetable (cf. “Hydrophobic Treatment of a Sampling Pipette”). – Fine aluminium capsules with a capacity of 30–40 mL. – Drying oven with ventilation, regulated at 105°C. – Thermometer. – Balances, range: 120 g, sensitivity: 0.1 mg. – Sets of two sieves with 0.2 and 0.05 mm mesh, with a vibrating sieving machine. – Rotary agitator able to receive 10 or 20 cylinders, (30 rpm). – Standard tributylchlorosilane. – Standard 1-chloronaphtalene. Preparation of Pipette Hydrophobic Treatment of a Sampling Pipette This treatment (Walker method) makes the walls of the pipette nonwetable and eliminates the need for rinsing between sampling. Prepare the smallest possible quantity of 4% solution of tributylchlorosilane in chloronaphtalene. Clean, carefully degrease the pipette, dry it well then treat the inside of the pipette by aspiration; drain, leave to dry for a minimum of 24 h at room temperature. Calibration of the Pipette – Overload Reagent This calibration should be done periodically. The same dispersing liquid used for the analyses is used again but the temperature should be checked as it influences viscosity. Put five fractions of 20 mL in tarred capsules, then weigh (± 1/10 mg). The average weight corresponds to the volume

36

Mineralogical Analysis

removed. Dry for 5 h at maximum 105°C, weigh the capsules (± 1/10 mg) to determine overload due to the reagent. Procedure All operations should be carried out at 20°C in an air-conditioned room, and equipment and reagents should be kept at the same temperature. Control of Dispersion After agitation, lay the sedimentation cylinders out in line on the lab table, and open, taking care to include any deposits on the stopper.

Fig. 2.5. Sampling system with the Robinson pipette (on the right) and the Andreasen pipette (on the left)

First check the state of the suspension. Carefully check whether flocculation has occurred although this may be only partial and impossible to see. In case of doubt, complete the following steps: measure the temperature of the suspension, calculate (2.1) the falling time at 5 cm and 10 cm (for particles of 0.02 mm), hand shake the cylinder by turning it upside down and back for 1 min; start the chronometer, lower the pipette to 5 cm. Ten

Particle Size Analysis

37

seconds before the time is up, remove a 10 mL sample and place it in a capsule. Lower the pipette to 10 cm and remove a sample at the corresponding time under the same conditions. Dry the capsules containing the samples in the drying oven at 105°C for 24 h, then weigh them (± 1 mg); the two weights should be identical, any difference indicates flocculation and its extent. If any doubt remains about possible flocculation, it is better to start the analysis again. Certain apparatus make it possible to identify the optimal zone of dispersion for certain clays by bringing into play their property of orientation in an electric field. First Sampling (Clay + Silts) In the case of big series, a team of two people can be used, who should respect the timing very strictly (Table 2.4). Hand shake the cylinder containing the 1,000 mL suspension by turning it upside down and back to put all the deposits in suspension then place it on the counter top and start the chronometer (remove the stopper after the particles have been removed by agitation). – Close the three-way stopcock of the pipette; lower the pipette until the tip touches the surface of the suspension. Note the position of the index on the scale. Approximately 30 s before the time is up, (4 min 48 s at 20°C for 10 cm, see Table 2.5), carefully lower the pipette in the first agitated cylinder to the selected depth (here 10 cm). – Exactly 10 s before the sampling time, begin aspiration of 20 mL (speed intake 1 mL s–1) by slowly opening the stopcock. The distribution around the exact sampling time gives the “average”’. When the liquid rises above the top of the stopcock, turn it off and run the overflow off through the side nozzle.
Table 2.4. Procedure for stirring a series of sedimentation cylinders

stirring first cylinder 1 min

1 min

stirring second cylinder 1 min

1 min

etc.

when stirring is finished, start the lag chronometer: this is the time beginning of the timed period

lag time

Remove the pipette, quickly wipe the outside of the tube, and empty its contents into a previously tared 50 mL weighing bottle. Evaporate to dry and dry in the drying oven at maximum 105°C for 3 h. Weigh dry residue to precisely 1/10 mg. Correct the weight for overload due to the reagent.

38

Mineralogical Analysis

Remarks Aspiration should not be too fast to limit turbulence and to avoid aspirating particles with a larger diameter than those of the selected sampling range. Aspiration must also be regular. The pipette does not need to be rinsed between sampling since it has received hydrophobic treatment (cf. “Hydrophobic Treatment of a Sampling Pipette”) and any error due to retention is negligible compared with other causes of error. Blank assays should be made with the dispersant alone (cf. “Calibration of the Pipette – Overload Reagent”). Weigh the sample and the blank after 24 h in the drying oven at 105°C. Second Sampling: Clay
Table 2.5. right: Sampling time of particles (d = 2.65) by sedimentometry with a Robinson-Köhn pipette at a depth of 10 cm. left: Sampling depth of the clay-size fraction at different times clays < 2 µm depth of sampling in cm after 5h 6h 7h 8h 6.2 6.4 6.5 6.7 6.9 7.0 7.5 7.7 7.9 8.1 8.3 8.5 8.8 9.,0 9.2 9.4 9.7 9.9 10.0 10.3 10.5 10.8 11.0 11.3 temperature T (°c) clays <2 µm < 5 µm < 20 µm < 50 µm

falling time at 10 cm

20 21 22 23 24 25

8 h 00 min 7 h 48 min 7 h 37 min 7 h 26 min 7 h 16 min 7 h 06 min

1 h 16 0 h 04 0 min 47 s min 48 s min 48 s 1 h 15 0 h 04 sedimentamin 00 s min 41 s tion 1 h 13 0 h 04 min 12 s min 34 s 1 h 11 0 h 04 min 30 s min 28 s 1 h 09 0 h 04 min 54 s min 22 s 1 h 08 0 h 04 min 18 s min 15 s method not possible

Proceed as above after 8 h of sedimentation at 20°C (if necessary, a smaller depth can be used to sample clay on the same day: for example, 7 h at 8.8 cm at 20°C – see Table 2.5). Weigh the residue exactly (± 1/10 mg) and correct the weight for overload due to the reagent (blank).

Particle Size Analysis

39

Intermediate sampling can be performed; remember to take suspensions containing the coarsest phases first and the finest last. For this it is better to use an Andreasen pipette. Washing and Measuring Fine and Coarse Sands After the last sampling of clays, siphon off the supernatant liquid to 5 cm from the bottom; decant the deposit in 1,000 mL beakers. Add deionized water with a little hexametaphosphate (approximately 700–800 mL); agitate vigorously to put the deposit in suspension; after the time necessary for 0.02 mm particles to fall below the limit of aspiration of the siphon, siphon off the supernatant liquid (Fig. 2.6).

Fig. 2.6. Washing of sands by decantation: left, first siphon off, right, continue siphoning until supernatant is clear (A = falling height of 0.02 mm particles)

Again add water with a little hexametaphosphate; continue washing until the liquid supernatant is clear; finish with one final washing with distilled water; eliminate the maximum amount of water possible per decantation, quantitatively decant the deposit of sands in a capsule, put to dry in a ventilated drying oven at 105°C; after cooling, weigh total sands; put the sands on the top of two superimposed sieves, one with a 0.2 mm mesh (AFNOR 24), the other with a 0.05 mm mesh (AFNOR 18); sieving with a vibrating apparatus must be complete; check there are no cemented aggregates or plant debris. The mesh of the 0.2 mm sieve represents coarse sands, and the mesh of the 0.05 mm sieve represents fine sands. The coarse silts or very fine sands are determined by calculating the difference between total sands and the sum of coarse sands and fine sands.

40

Mineralogical Analysis

Remark The washing–decantation operations are long and tiresome, particularly as many samples are required in routine analyses. It is possible and advisable to use an automated system to wash the sands, e.g. the one developed by Susini (1978). Causes of Error A strict procedure is required to ensure the temperature remains stable for the duration of sedimentation. This can be achieved by immersing the cylinders in a thermostat bath, but this device is not really suitable for the treatment of the large series required in many laboratories. Consequently the cylinders are simply laid out in line on the lab table. The temperature should be taken in one of the cylinders at the beginning. Since the liquid medium presents good thermal inertia, variation is not very great over a short period, i.e. for the first sampling of about 4 – 5 min. The most favourable temperature is between 15°C and 25°C; above 30°C there is a risk of flocculation; below 15°C the times needed for sedimentation will be too long. For these reasons, in the absence of a thermostat, it is best to work in an air-conditioned room. But the main causes of error are: – Too abrupt entry of the pipette in the suspension. – Error in the depth of sampling. – Irregular or too rapid aspiration; certain authors recommend a time of 20 s per aspiration for a volume of 10 mL. These requirements exclude aspiration with the mouth. Optimal working conditions are guaranteed by making sure there is no variation between the way the operators perform the series of operations, i.e. moving the pipette, lowering the pipette into the suspension, exact timing of the beginning of the sampling, very regular sampling, exact volume, careful removal of the pipette, draining of the sample into the capsule, next sampling. The ideal solution is to use a simple automatic unit like that described in Pansu et al. (2001). However, even if the traditional manual system has to be used it is preferable to carry out aspiration with a small electric pump with a fixed flow; peristaltic pumps fulfil this function well.

Particle Size Analysis

41

Calculations Collected Data Mass of soil sample (air dried) = Moisture correction coefficient = mass sample after drying at 105°C /m = Volume of the sample = Mass of blank (reagents without sample) after drying 105°C = Mass of first sample (clay + silt) after drying 105°C = Mass of second sample (clay) after drying 105°C = Mass of total sands after drying 105°C = Mass of coarse sands (rejected by 0.2 mm sieve) after drying at 105°C = Mass of fine sands (rejected by 0.05 mm sieve) after drying at 105°C = Calculation of the Results in% of Soil Dried at 105°C Clays = C = (m2 – mB) × 1000 × 100 / (Vp × m × K) Silts = Si = (m1 – m2 × 1000 × 100 / (Vp × m × K) Fine sands = FS = 100 × m5 / (m × K) Coarse sands = CS = 100 × m4 / (m × K) Total sands = S = 100 × m3 / (m × K) Coarse silts = CSi S – (FS + CS) If limestone is present (Table 2.3, B), and particle size analysis was performed without destruction of carbonates, carbonates should be determined on the separated fractions which provides information about the distribution of limestone. Checking and Correction of the Results Taking into account the moisture correction factor (cf. “Calculations” under Sect. 2.2.3), the sum: clays + silts + total sands + organic matter + if necessary, carbonates, soluble salts, gypsum, must be between 95 and 102%, preferably between 98 and 102%. Soils rich in organic matter can provide too high balances: in the event of incomplete destruction during pretreatment, organic matter can be counted twice. A too small sum results from losses during the pretreatments. In the majority of cases, it is impossible to determine the exact proportions of losses in organic matter, soluble salts, carbonates and gypsum. An overall estimate of the losses can be made as follows: using the cylinder in which the samplings were made, add 10 mL of 1 mol (CaCl2) L–1 solution and 1 mL of 1 mol (HCl) L–1 to flocculate colloids and prevent the formation of m K Vp mB m1 m2 m3 m4 m5

42

Mineralogical Analysis

calcium carbonate during drying in the drying oven. Allow the particles to deposit, completely remove the clear solution, put the deposit in a tared capsule, dry in the drying oven at 105°C, and then weigh. This gives mass mr from which losses during the treatments (organic matter, etc.) can be determined and the balances corrected. 2.2.4 Density Method with Variable Depth Principle This type of analysis is advantageous because it avoids fractionation of the sediment in dimensional classes and allows the construction of curves of distribution. In the density method proposed by Bouyoucos (1927, 1935, 1962), the heterogeneous suspension is considered to behave in the same way as a homogeneous liquid with the same density. Casagrande (1934) showed that it is then acceptable to use a float densimeter to measure the average density in the suspension column with the float. The plan of average density is located at a distance HR from the highest level of the suspension. From (9.4) and (9.4’) it separates the particles from the radius:

r = C −1/ 2 H r t

( )

1/ 2

,

(t = sedimentation time), and allows particle size analysis. The density method can be performed with permanent immersion, which fulfils the conditions of continuous measurement, or by temporary immersion, which resembles discontinuous measurement described in Chap. 1, without the same precision, but for routine measurements it has the advantage of avoiding sampling and weighing. On the other hand, the density method requires a larger number of samples than the pipette method. Equipment and Reagents – Cylinders identical to those described in “Equipment” under Sect. 2.2.3, special conical hydrometers to avoid accumulation of the particles on the surface, thermometer, pycnometers, magnifier with long effective focal spot. – Dispersing reagents of “Equipment and Reagents”, isoamylic alcohol.

Particle Size Analysis

43

Checking the Hydrometer Calculation of Falling Height of the Particles The relation of Casagrande makes it possible to calculate the value Hr giving the effective depth selected as falling height of the particles. For this calculation, the measurements shown on the hydrometer in Fig. 2.7 are required. The volume of the hydrometer is obtained by liquid displacement. Hr = h1 + 0.5 (h – V/S) (2.5)

V: volume in cm3 S: section in cm2 A graph can be drawn representing depths Hr as a function of the densities. This makes it possible to draw up the table giving the size of the particles as a function of temperature, time, and depth, for an unknown hydrometer. Checking the Graduations on the Hydrometer This test is made with pure water and a 2% solution of barium nitrate or chloride. Use a pycnometer to measure the density of water (note the temperature) then the density of the 2% solution (4 significant decimals). Take the same measurements by submerging the hydrometer in the test-tube successively containing the two liquids; read the densities with the help of a magnifier with a long focal spot (5 or 6 cm). Note the differences in the measurements obtained with the pycnometer and the hydrometer. If the relative values are the same, the hydrometer is valid even if the indications are not exact, because in the calculations the differences in density are used. If the differences between the pycnometer and hydrometer measurements are not constant, an abacus of transposition has to be established: values read on the hydrometer/actual values (this is very seldom the case). This method is similar to that described in Sect. 2.2.3 using 30 g of fine earth (soil air dried and sieved to 2 mm). For dispersion, add 30 mL of 102 g L–1 hexametaphosphate solution. Agitate for 4 h by upside down and back rotary shaker, transfer in the cylinders and complete to 1,000 mL.

44

Mineralogical Analysis

Fig. 2.7. Relation between density and falling height of the particles Hr (2.5) for a hydrometer with the following characteristics: V = 45 cm3, S = 28.26 cm2, h = 17 cm, h1 = 15.2 cm).

Procedure Preparation of the Sample

Fig. 2.8. Positioning of hydrometer

Particle Size Analysis

45

Measurement of the Clay + Silt (0.02 mm) Fraction
Table 2.6. Sizes of the particles as a function of temperature, time and the depth: valid for a standard hydrometer (laboratory of sedimentary sequences, IRD Bondy, France, unpublished data)
Particle Size t (°C) density→ (min ↓) 4 15 6 8 9 20 µm 4 6 20 8 9 (clays + silts) 25 3 4 6 8 particle size t (°C) density→ (hours ↓) 6 15 2 µm 8 24 6 20 (clays) 8 24 6 25 8 24 2.2 µm 1.9 1.1 2.1 1.8 1 2 1.4 1 2.6 µm 2.2 1.3 2.4 2.1 1.2 2.3 2 1.1 2.9 µm 2.5 1.4 2.7 2.3 1.4 2.6 2.2 1.2 3.1 µm 2.7 1.58 2.9 2.5 1.5 2.8 2.4 1.4 3.4 µm 2.9 1.7 3.2 2.7 1.6 3 2.6 1.5 1.020 8.7 cm 21 µm 17.4 15 14.2 20 16.4 14.2 13.4 21.8 19.9 15.4 13,4 1.020 1.015 11.6 cm 24.7 µm 20.1 17.4 16.4 23 19 16.3 15.4 25.2 21.8 17.8 15,4 1.015 1.010 14.4 cm 27.4 µm 22.4 19.4 18.2 25.8 21 18.2 17.2 28.1 24.3 19.8 17,2 1.010 1.005 17.2 cm 30.1 µm 24.5 21 20 28.2 23 19.8 18.8 30.7 26.6 21.6 18,8 1.005 1.000 20.1 cm 32.5 µm 26.5 23 21.6 30.3 24.5 21.4 20.2 33.2 28.7 23.4 20,3 1.000

If possible measurements should be made in an air-conditioned room at a constant temperature of 20°C. The cylinders containing the suspensions are grouped; check their volume has been completed to 1,000 mL, measure the temperature by referring to a table (such as Table 2.6) giving

46

Mineralogical Analysis

times of sedimentation for measurements at the selected temperature (for example 4–6–8–9 minutes). Hand shake the first cylinder by it turning upside down and back for 1 min, add an isoamyl alcohol drop anti-foamer, start timing, introduce the hydrometer very gently so as not to disturb the suspension (this is the most delicate part and a significant cause of error), take care that the hydrometer is maintained in the centre of the suspension, (if need be, make a paper guide see Fig. 2.8), take the readings at the top of the meniscus at 4–6–8 and 9 min. Continue in the same way with the following cylinder. Take a reading of the blank in a cylinder containing only the dispersant. Measurement of Clay Fraction (< 0.002 mm) The time counted starts at the beginning of agitation of the first cylinder during the first reading (clay + silt). To define this time, the average temperature has to be calculated from the beginning of sedimentation until the reading; take readings with the hydrometer at 6–8–24 h carefully respecting the 2 min interval used for the first sampling; transfer the hydrometer very carefully from one cylinder to another in order to avoid disturbances.

Fig.2.9.

Example of abacus: sizes of particles are a function of the time of sedimentation for a given hydrometer and a density close to 1.010.

Remarks As can be seen in Table 2.6, that selected times do not correspond exactly to the selected particle size; however, in a small space of time, variation in particle size is considered to be continuous; this makes it possible to plot an abacus around a given depth (Fig. 2.9) which then makes it possible to define an exact falling time for a given particle size.

Particle Size Analysis

47

Calculations The percentage of particles P corresponding to a given density is obtained from the equation:
P= m ρs − ρ f

ρs = density of solid = 2.65 for soil; ρf = density of the dispersing solution; in our conditions at t°C, one can calculate ρf by means of (2.2); ρ = density read at t°C; δρ = corrections + or – on readings to bring them to 20°C (Table 2.6);
V = volume, 1,000 mL; m = soil sample, 30 g; The calculation can be simplified by calculating K = 100 ρs / m (ρs – ρf ) which gives P% = K [(ρ ± δρ) – ρf ] V with m = 30 g, at 20°, it gives K = 5.35 Determination of Sands Use the same technique as that described in Sect. 2.2.3. 2.2.5 Density Method with Constant Depth The density method with variable depth (cf. 2.2.4) has the advantage of greater speed compared to the pipette method as well as allowing uninterrupted measurements if required. However, as the depth of immersion is not constant, the degree of precision is lower and calculations are longer. The chain hydrometer (de Leenheer system) makes it possible to measure the density of the suspension to a given constant depth of approximately 20 cm, which in turn, makes it possible to approach the principle and the precision of the pipette method while avoiding sampling and weighing (De Leenheer and Macs 1952; De Leenheer et al. 1955; De Leenheer and van Hove 1956; van Ruymbeke and de Leenheer, 1954). The apparatus (Fig. 2.10) is composed of an immersion body at the end of an arm with (1) a pointer that identifies the level of the liquid and (2) at the top, a support that can receive overload weights in the form of riders, and an equilibrium chain that allows a very fine fit when the pointer locates the surface of the liquid. The depth of sedimentation is represented by the distance from the point of the needle located in the middle of the body of the hydrometer.

(

100 ρ s

)

(( ρ

±δ ρ − ρ f

)

)V

(2.6)

48

Mineralogical Analysis

Fig. 2.10. Principle of chain hydrometer after De Leenheer and van Hove(1956).

One minute before measuring time, carefully introduce the hydrometer into the suspension; add the weights in such way that the reference mark of the needle is 1 – 2 cm above the level of the liquid. At the precise time of the sampling, adjust the needle so that it is in contact with the liquid by quickly adjusting the chain with the screw device. The chain can cause an overload of 100 mg. The reading for total overload gives the weight of the hydrometer. Having determined in advance the volume of the hydrometer (by immersion in distilled water), one thus obtains: density (ρ) = weight of the hydrometer / hydrometer volume Continue the calculations in the same way as for the density method with variable depth (2.6). 2.2.6 Particle Size Analysis of Sands Only Place 100 g of fine earth (standard 2 mm preparation from a perfectly homogenized batch) in a 1,000 mL beaker with 100 mL hydrogen peroxide brought to 30 volumes; leave to act in a cold place overnight, then transfer on a moderately heated hotplate; add hydrogen peroxide in small fractions until complete destruction of the organic matter, then eliminate excess hydrogen peroxide by boiling without going to dry.

Particle Size Analysis

49

Table 2.7. The two sieving columns used successively for particle size analysis of sands column 1 mm 2 1.6 1.25 1 0.8 0.63 0.50 0.40 AFNOR standard 34 33 32 31 30 29 28 27 mm 0.315 0.25 0.20 0.16 0.125 0.100 0.08 0.063 0.050 column 2 AFNOR standard 26 25 24 23 22 21 20 19 18

Transfer the residue in a cylinder and add 500 mL distilled water and 25 mL of 52 g L–1 sodium hexametaphosphate solution. Place in a rotating shaker for 4 h in the same way as for complete particle size analysis. After this operation, transfer the content of the cylinder on a 0.05 mm mesh sieve (French standard AFNOR NF-X-11-504 module°18); wash the residue under running water. Place the well-washed sands of the 0.05 mm sieve in a 250 mL beaker. Add 100 mL of 6 mol (HCl) L–1 solution, cover with a beaker cover, and boil gently for 2 h to dissolve iron. After cooling, decant and wash by successive decantation until complete elimination of the acid; transfer again on a 0.05 mm sieve, wash, and transfer quantitatively the sands in a capsule; dry for 24 h in drying oven at 105°C. Let cool and weigh total sands. Sieve dry total sands successively on the two columns of sieves (Table 2.7). Place the columns successively on a vibrating sieve machine (Pansu et al., 2001). Transfer in column 2 the fraction collected at the bottom of column 1. Sieve 10 min and weigh each fraction (± 0.01 g). Checking: sum of weighings of each fraction = total sands.

50

Mineralogical Analysis

2.3. Automated Equipment
2.3.1 Introduction The phenomena of slaking of the soils requires precise knowledge of the grain-size distribution of the 2–20 µm fraction; sediment studies require a distribution of the fine phases down to 0.1 µm or even lower. In agronomy, the horizons comprising the formation of clay are studied using ratios for “coarse clay < 2 µm/fine clay < 0.2 µm”. Since gravity methods cannot provide all the answers, a range of different techniques is required. Table 2.1 summarizes the main methods used for measurement of the particle-size distribution of soils. Some methods are well suited for the repetitive measurements needed for studies in the fields of pedology, agronomy, geology or sedimentology; others are more suitable for detailed and in-depth studies. The choice of a method will depend on: − the degree of precision required, reproducibility and repeatability, and a good correlation with the pipette reference method (cf. 2.2.3) despite its defects; − the extent of the particle size field and possibility of extending it to sub-micronic or nanometric particles; − the speed of execution, the time needed to produce a result, the flexibility of use, the possibility to significantly increase the number of analyses; − the possibility of recovering the particle fractions for later measurements, or of taking other measurements simultaneously (continuous analyses, etc.); − the cost of equipment and personnel, the importance of the request and available space; − continuous data acquisition and exploitation (monitoring, calculations, histograms and cumulative frequency curves). Given the wide granular spectrum of soils, combinations of individual methods that do not cover the complete spectrum are often used. There should be a significant overlap between the methods. The pipette method remains the reference method for all comparisons; analysis by laser diffraction makes it possible to extend the spectrum to the sub-micronic field but still identify silts. The true representativeness of equivalent diameters in a given class can be checked using a microscopic method and image analysis.

Particle Size Analysis

51

2.3.2 Methods Using Sedimentation by Simple Gravity The majority of apparatus designed to mechanize the pipette method are not widely distributed, as each laboratory tends to develop devices suitable for its own needs. For example, the automatic particlemeasurement instrument distributed by “Technology Diffusion France” (Pansu et al., 2001) is based on sedimentation in a thermostated cupboard with an automated pipette. Analysts, especially those in the industrial sector, need methods that obtain rapid results with good repeatability and the possibility of calibrating the apparatus (using calibrated microball powders). For the main types of automated equipment available and the names of manufacturers see Pansu et al. (2001). Sedimentation Balances and Automated Sieve Machines for Wet Measurements These balances (Sartorius, Cahn, Mettler, etc.) make it possible to continuously record the process of sedimentation between approximately 1 and 150 µm (Fig. 2.11), the higher fractions being in the domain of automated wet or dry sieve test machines (Micromeritics, Seishin, etc.).

Fig. 2.11. Diagram of a particle size balance with constant equilibration .

Sedimentation is carried out after dispersion on samples of reduced weight, i.e. about 1–2 g, in a thermostatic enclosure. Continuous automatic recording of the weight of the sediments deposited makes it possible to create cumulative mass curves as a function of time. Depending on the degree of automation and calculation, frequency charts can also be created. The measurements are reproducible, but require a long time to perform and are thus not suitable for series analysis.

52

Mineralogical Analysis

Systems Using Simple Gravity and Measurement by X-ray The particles are prepared and put in suspension (as described in sections X-ray beam). The particles absorb a quantity of X-ray proportional to their number. The resulting intensity is measured by a scintillation counter. At the beginning, the resulting intensity of X-rays is at a minimum, then the falling particles cause an increase in the intensity transmitted. To reduce measurement time, the cell containing the suspension gradually moves downwards and the fixed X-ray beam sweeps a portion of the suspension increasingly close to the surface. All these movements are controlled by computer, and the position of the cell is a logarithmic function of time, coupled with the x-axis of the recorder, which makes it possible to determine the diameter that corresponds to the position of the cell. The smallest diameter it is possible to measure is 0.1 µm and the largest is 100 – 300 µm, depending on the model. Continuous measurements make it possible to express the results in the form of cumulative curves of histograms of weight, or number of surface particles, etc. Computer interfaces make it possible to store the results of the analyses, and a sampler equipped with a carrousel allows uninterrupted treatment of several samples. It should be noted that X-rays with wavelengths of less than 10 nm are well suited for the measurement of particles which would be impossible to measure in visible light (100 – 800 nm i.e. similar to the diameter of fine clay particles). Repeatability is satisfactory and measurement up to 2 µm takes about 10 min. This type of equipment has been the subject of comparative studies with the pipette method for the analyses of soils (e.g. Delaune et al., 1991). It is useful for the 50–1 µm fraction, but needs a longer time for finer particles (< 0.2 µm in 50 min). However, undervaluation of coarse silts has been observed when they comprise more than 20% of the soil. System Using Simple Gravity and Measurement by Light Absorption or Scattering Methods based on photo-sedimentation (nephelometry, turbidimetry) are subject to many interferences. Their reproducibility and repeatability are low due to the use of white light or monochromatic light in the visible spectrum. These methods should only be used for rapid comparisons within homogeneous families.

Particle Size Analysis

53

Methods Using Elutriation In these methods, the falling speed of the particles (the mobile solid phase which is easy to measure) is lower, equal or higher than the speed of the fluid (non-stationary mobile liquid phase). The liquid phase circulates in the reverse direction to the particles, making it possible to sort them; the finest particles migrate upwards or fall by gravity. The different fractions can be recovered. This system is suitable for certain studies on sediments, but measurements take a long time and are not really suitable for repetitive analysis. 2.3.3 Methods Using Accelerated Sedimentation Principle In practice, methods using simple gravity cannot be used for particle sizes <2 µm because of the extremely long time needed for sedimentation of the finest particles. It is difficult to maintain the cylinders of sedimentation without convection currents for long periods and to withdraw the particles from the Brownian movement. However, acceleration of gravity by centrifugation makes it possible to exceed the limits and to mitigate the effect of Brownian movement. The techniques of separation and the recovery of granulometric phases by this process are discussed in Chap. 3. Apparatus Using Centrifugal Discs Some equipment uses successively first simple gravity with the vertical rotor remaining stationary for the largest particles, and second gravity with centrifugation at speeds of 1,800 to 8,000g (Horiba, Shimadzu, Seishin, Union-Giken, Joyce-Loebl-Vickers, etc.). Analysis is continuous, and recording makes it possible to automatically create curves and histograms. Masses of soil of the order of 1 g can be treated in this way. Depending on the manufacturer, the measuring cells are intersected either by a filtered incandescent light with measurement of absorption, or – very exceptionally – by laser or X-ray detection (Brook Haven). Particles of 0.01 µm to 100 µm can be identified. Other manufacturers use horizontal discs and samplings at a given distance and at a given time. The fractions are dried, weighed and possibly subjected to other analyses (Fritsch, Simcar, Joyce-Loebl, etc.).

54

Mineralogical Analysis

The performances of these apparatus are not always equal for series analysis, and repeatability is not always within the range usually obtained with the pipette method. For sub-micrometric analysis, one manufacturer offers an ultracentrifuge with a titanium disc at 100,000g in a partial vacuum to avoid heating and noise. A UV scanner (280 nm) makes it possible to analyze soil particles of less than 500 nm (Beckmann Spinco). Micro-methods using Field flow fractionation (FFF) are still not reliable enough for widespread use in soil studies. 2.3.4 Methods Using Laser Scattering and Diffraction Laser particle-measurement instruments have undergone spectacular development and can now be used for an increasing range of particle sizes. Certain equipment make it possible to cover ranges from 0.1 to 2,000 µm, but in general, apparatus are particularly powerful for a more limited range. One range is dedicated to sub-micronic, or even nanometric particles, while others with a wider range are particularly useful for the particle size analysis most usually required by soil laboratories. These measurements are not based on sedimentation and must consequently be calibrated. A dispersing liquid containing suspended particles circulates in a measuring cell intersected by a monochromatic Laser beam collimated by a condenser on a window of analysis of a defined surface. The light of the Laser is diffracted on the outside of the particles and the angles of diffraction are inversely proportional to the size of the particles. An optical system collects the signals which are analyzed by Fourier transformation and discriminated on a detector engraved with predetermined angles. The signal is treated to extract the distribution of the particles. The results can be expressed in the form of curves: by average diameter (particle size distribution) expressed as a percentage of total weights, by histograms of weight, surface, number of particles, volume, etc. 32 – 64 classes of sizes can be measured (Malvern, Cilas-Alcatel, Coulter, etc.). Certain apparatus allow either proportioning on a suspension, or on dry powder, which can be useful for analyzing silts. Serial deflocculation on line is possible by ultrasound. Loading 40 samples with a sample distributor and a using distributor for reagents makes it possible to work without continuous monitoring. Analytical files enable methods to be pre-determined, including the dispersants. The procedures are simple but vary considerably with the

Particle Size Analysis

55

apparatus and it is consequently impossible to give a detailed procedure here. 2.3.5 Methods Using Optical and Electric Properties Analysis of the distribution of sub-micron particles (3 nm to 3 µm) combines measurements of pH, temperature, conductivity, and relative viscosity making it possible to control the stability of a suspension and the electro-kinetic potential (Zeta potential, potential difference between the dispersed surface layer and the medium of dispersion). The ionic force is measured in an electrophoresis quartz tank with a Pt–Mo electrode on particles measuring from 1 to 1,000 µm (Malvern, Brookhaven, Coulter, Micromeritics, Mono-Research Lab., Zetameter Inc., Matec Applied Science, etc.). The effective surface charge of the particles is determined by the measurement of mobility in a liquid–solid system, the permittivity of the liquid being known. This enables the study of the phenomena of flocculation and dispersion. Other apparatus are designed for the study and optimization of the dispersion of certain clays for industrial use. They make it possible to differentiate flocculated and deflocculated particles. For example, primary Kaolinite particles consist of regular hexagonal discs. The nature of these particles means that in the presence of an electric field a dipole is induced, causing alignment with the field. The neo-aggregates formed by flocculation consist of clusters of randomly arranged primary particles, out of alignment with the electric field. To measure the relative proportions of flocculated and dispersed particles, the suspension is intersected by a Laser beam and the diffused light is analyzed. The result is quantified. Measurements made in different conditions (dispersing concentration, the nature of the dispersant, pH, etc.) enable optimization of the analyses. 2.3.6 Methods Allowing Direct Observation of the Particles

Optical and Electronic Microscopy – Radiation Counter and Image Analyzer These direct methods are based on the use of optical microscopy, or possibly of electronic microscopy (cf. Chap. 8). Particle fractions isolated by gravity can be used among others. In electronic microscopy, the preparations must be dried and presented on grids or plates (MET-MEB).

56

Mineralogical Analysis

Microscopy makes it possible to directly observe the population of particles and their morphological parameters. Shape, length, width, thickness or diameter can be defined on micro-samples, and the fractal properties estimated. Comparison of the particles of the same fraction makes it possible to judge the quality of fractionation (use of tests of bulky diatoms, regularity, influence of density, etc.). Counting can be done by a radiation counter, or by an image and texture analyzer. In electronic microscopy with an EDX probe it is possible to perform chemical analyses; and in optical microscopy, to use infra-red radiation. This equipment offers a wide range of possibilities (Quantachrome, Zeiss, Leitz, etc.) and makes it possible to establish percentages of cumulated mass, distribution by size of particles, in terms of number, mass and specific surface area (0.1 – 300 µm). 2.3.7 Methods Using Conductivity While passing through a gauged opening, a particle displaces a volume of electrolyte which modifies electric resistance (differential conductivity). This change in resistance is a function of volume. Counting allows particles to be grouped in classes using an amplitude discriminator. The results are expressed in 16 counter channels as total percentage weight or the number of particles of a specific dimension; results can be presented in the form of graphs or tables. The apparatus based on this principle are counters that make it possible to ignore density which can be useful when dealing with soils rich in iron oxides with a high density (4 – 5), i.e. well above the mean of 2.65 used for sedimentation by simple gravity. The shape of the particles is significant for the accuracy of the measurements. Measurements are possible between 1 and 5,000 µm (Coulter), but the optimal field of measurement depends on the choice of a suitable opening. Particles of less than 1 µm are often underestimated. The preparation of the samples is identical to the standard method (cf. Sects. 2.2.1 and 2.2.2).

References
Atterberg A (1912) Die mechanische Bodenanalyse und die klassifikation der mineral böden sechwedens. Int. Mitt. Bodenk., 2, 312–342 Baize D (2000) Guide des anal yses courantes en pédologie., INRA, France, 257 p Bouyoucos GS (1927) The hydrometer as a new method for mechanical analysis of soils. J. Soil Sci., 23, 343

Particle Size Analysis

57

Bouyoucos GS (1935) A hydrometer method for making mechanical analysis of soils. Bull. Am. Ceram. Soc., 14, 259 Bouyoucos GS (1962) Hydrometer method improved for making particle size analysis of soils. Agron. J., 54, 464–465 Casagrande A (1934) Die Aräometer-methode zur Bestimmun der kornverteilung von Boden und anderen materialen. Springer J. De Leenheer L and Van Hove J (1956) Werkwijze voor de mechanische analyse met de kettinghydrometer. Rijksland Bouwhogeschool (Gand)., XXI, 249–274 De Leenheer L and Maes L (1952) Analyse granulométrique avec l’hydromètre à chaîne. Bull. Soc. Belge de Géologie., 61, 138–164 De Leenheer L, Van Ruymbeke M and Maes L (1955) L’analyse mécanique au moyen de l’hydromètre à chaîne. Silicates Industriels., Tome XX, n° 6– 7, 1–7 Delaune M, Reiffsteck M and Feller C (1991) L’analyse granulométrique de sols et sédiments à l’aide du microgranulomètre sédigraph 5000 et comparaison avec la méthode à la pipette Robinson. Cahiers ORSTOM sér. Pédol., 26, 183–189 Gee GW and Bauder JW (1986) Particle-size analysis. In Methods of Soil Analysis. Part 1 Physical and Mineralogical Methods., Klute A. Ed. Chap. 15. American Society of Agronomy. Soil Sci. Soc. Am., 383–411 Gras R (1988) Physique du sol pour l’aménagement, Masson, Paris, 587 p Hénin S (1976) Cours de physique du sol, vol. 1. Orstom-Editest, Bruxelles, 159 p Jackson (1969) Soil Chemical Analysis – Advanced Course., 2nd ed. University of Wisconsin, Madison, WI Mériaux S (1954) Contribution à l’étude de l’analyse granulométrique. Ann. Agro., I, 5–53, II, 149–205 Pansu M, Gautheyrou J and Loyer JY (2001) Soil Analysis – Sampling, Instrumentation and Quality Control, Balkema publishers, Lisse, Abington, Exton, Tokyo, 512 p Pétard J (1993) Les méthodes d’analyse. T1 Analyse de sols., Notes techniques laboratoires communs d’analyse, Orstom, Nouméa, Paris Rouiller J, Burtin G and Souchier B (1972) La dispersion des sols dans l’analyse granulométrique. Méthode utilisant les résines échangeuses d’ions. Bull. ENSAIA, Nancy, France, 14, 183–204 Stokes GG (1851) On the effect of the lateral friction of fluids on the motion of pendulums. Trans. Cambridge Phil. Soc., 9, 8–106 Susini J (1978) Realisation d’un ensemble automatique de lavage des sables de l’analyse granulométrique. Cah. ORSTOM Série Pédol., 16, 339– 344 Tamm O (1922) Eine Methode zur Bestian on vag der anorganischen komponenten des Gelkomplexes in Boden. Meddel. Staters Skogsfïrsöksanst (Suède), 19, 385–404 Van Ruymbeke M and De Leenheer L (1954) Etude comparative d’analyses granulométriques par décantations successives et par l’hydromère à chaîne. Actes et C. R. du Vème Congrès International de la Science du Sol (Leopoldville), II, 322–328

58

Mineralogical Analysis

Vieillefon J (1979) Contribution à l’amélioration de l’étude analytique des sols gypseux. Cah. ORSTOM Sér. Pédol., XVII, 195–223 X 31-107 (1983) Analyse granulométrique par sédimentation. Méthode de la pipette. In Qualité des sols 3°ed., AFNOR, 357–371

Bibliography
Generality
Barth HG and Shao-Tang Sun (1991) Particle size analysis. Anal. Chem., 63, 1R– 10R. Chamayou H and Legros JP (1989) Les bases physiques, chimiques et minéralogiques de la Science du sol, Tech. Vivantes, ACCT Presses Univ. de France, 593 p. Guillet B and Rouiller J (1979) La granulométrie. In Pédologie, constituants et propriétés des sols, Bonneau and Souchier ed., Masson, 317–321. Johnston, Farina MPW and Lawrence JY (1987) Estimation of soil texture from sample density. Commun. Soil Sci. Plant Anal., 18, 1173–1180. Jones JL, Kay JJ, Park JJ and Bishop CK (1980) The determination of particle size distribution in soil. A collaborative study. J. Sci. Food Agric., 31, 724–729. Loveland PJ and Whalley WR (1991) Particle size analysis. In Soil Analysis: Physical Methods., Smith KA, Mullins CEJ ed., Dekker, New York 271–328. Smith RB and Pratt DN (1984) The variability in soil particle size test results by various sub sampling techniques. J. Soil Sci., 35, 23–26. Syvitski JPM (1991) Principles, methods and applications of particle size analysis. Cambridge Univ. Press., 366 pages.

Pre-treatment

Organic Matters
Douglas LA and Fiessinger F (1971) Degradation of clay minerals by H2O2 treatments to oxidize organic mater. Clays Clay Miner., 19, 67–68 Fisher WR (1984) The oxidation of sol organic matter by KBrO for particle size determination. Commun. Soil Sci. Plant Anal., 15, 1281–1284 Harada Y and Inoko A (1977) The oxidation products formed from soil organic matter by hydrogen peroxide treatment. Soil Sci. Plant Nutr., 23, 513– 521

Particle Size Analysis

59

Langeveld AD Van, Gaast SJ Van der and Eisma D (1978) A comparison of the effectiveness of eight methods for the removal of organic matter from clay. Clays Clay Miner., 26, 361–364 Lavkulich LM and Wiens JH (1970) Comparison of organic matter destruction by hydrogen peroxide and sodium hypochlorite and its effects on selected mineral constituents. Soil Sci. Soc. Am. Proc., 34, 755–758 Omueti JAI (1980) Sodium hypochlorite treatment for organic matter destruction in tropical soils of Nigeria. Soil Sci. Soc. Am. J., 44, 878–880 Sequi P and Aringhieri R (1977) Destruction of organic matter by hydrogen peroxide in the presence of pyrophosphate and its effect on soil specific surface area. Soil Sci. Soc. Am. J., 41, 340–342 Visser SA and Caillier M (1988) Observations on the dispersion and aggregation of clays by humic substances. I – Dispersive effects of humic acids. Geoderma, 42, 331–337 Vodyannitskii Yu N, Trukhin VT and Bagina OL (1989) The action of perhydral upon iron oxides in soil. Dokuchzer soil Sci. Inst. (Moscou), 1, 20–21

Eliminate Organo-Minerals Compounds
Harward ME, Theisen AA and Evans DD (1962) Effect of iron removal and dispersion methods on clay mineral identification by X-Ray difraction. Soil Sci. Soc. Am. Proc., 26, 535–541 Mehra OP and Jackson ML (1960) Iron oxide removal from soils and clays by a dithiomite-citrate system buffered with sodium bicarbonate. In Clays and Clay Minerals. Proc. Seventh Conf. Natl Acad. Sci. Natl Res. Counc. Pub., 237–317

Eliminate Soluble Salts – Gypsum
Rengasamy P (1983) Clay dispersion in relation to changes in the electrolyte composition of dialysed red–brown earth. J. Soil Sci., 34, 723–732 Rivers ED, Hallmark CT, West LT and Drees LR (1982) A technique for rapid removal of gypsum from soil samples. Soil Sci. Soc. Am. J., 46, 1338– 1340

Suspension – Dispersion – Flocculation
Balli P (1965) Critères de la qualité de la suspension en vue de l’analyse granulométrique. Science du sol, 1, 15 Bartoli F, Burtin G and Herbillon AJ (1991) Disaggregation and clay dispersion of oxisols: Na Resin, a recommended methodology. Geoderma, 49, 301–317 Brewster GR (1980) Effects of chemical pretreatment on X-Ray powder diffraction characteristics of clay minerals derived from volcanic ash. Clays Clay Miner., 28, 303–310 Colmet-Daage F, Gautheyrou J, Gautheyrou M, Kimpe C de (1972) Dispersion et étude des fractions fines des sols à allophane des Antilles et

60

Mineralogical Analysis

d’Amérique latine. 1ère partie: Techniques de dispersion. Cah. Orstom, Sér. Pédol., Vol. X(2), 169–191 Demolon A and Bastisse E (1935) Sur la dispersion des colloïdes argileux. Applications à leur extraction. Annales Agronomiques, 1–15 Dong A, Chesters G and Simsiman GV (1983) Soil dispersibility. Soil Sci., 136, 208–212 Egashira K (1981) Floculation of clay suspensions separated from soils of different soil type. Soil Sci. Plant Nutr., 27, 281–287 Forsyth P, Marcelja S, Mitchell DJ and Ninham BW (1978) Stability of clay dispersions. In Modidication of Soil Structure., Emerson, Bond, Dexter Ed. Wiley, New York. 2, 17–25 Goldberg S and Forster HS (1989) Floculation of reference clays and arid soil clays as affected by electrolyte concentration, exchangeable section percentage, sodium adsorption ratio, pH and clay mineralogy. Annual Meeting – Clay Minerals Society, 26, 35 Gupta RK, Bhumbla DK and Abrol IP (1984) Effect of sodicity, pH, organic matter and calcium carbonate on the dispersion behavior of soils. Soil Sci., 137, 245–251 Keren R (1991) Adsorbed sodium fraction’s effect on rheology of montmorillinite–kaolinite suspensions. Soil Sci. Soc. Am. J., 55, 376– 379 Manfredini T, Pellacani GC, Pozzi P and Corradi AB (1990) Monomeric and oligomeric phosphates as deflocculants of concentrated aqueous clay suspensions. Appl. Clay Sci., 5, 193–201 Miller WP, Frenkel H and Newman KD (1990) FLoculation concentration and sodium/calcium exchange of kaolinitic soil clays. Soil Sci. Soc. Am. J., 54, 346–351 Ohtsubo M and Ibaraki M (1991) Particle-size characterzation of flocs and sedimentation volume in electrolyte clay suspensions. Appl. Clay Sci., 6, 181–194 Oreshkin NG (1979) Device for tating suspension samples for the particle-size analysis of soils. Soviet Soil Sci., 4, 136-138 Reddy SR and Fogler HS (1981) Emulsion stability: determination from turbidity. J. Colloid Interface Sci., 79, 101–104 Reddy SR, Fogler HS (1981) Emulsion stability: delineation of different particle loss mechanisms. J. Colloid Interface Sci., 79, 105–113 Robinson GW (1933) The dispersion of soils in mechanical analysis. Bur. Soil Sci. Tech. Commun., 26, 27–28 Shaviv A, Ravina I and Zaslavsky P (1988) Floculation of clay suspensions by an anionic soil conditioner. Appl. Clay Sci., 3, 193–203

Ultrasonic Dispersion
Arustamyants YEI (1990) Optimizing the ultrasonic preparation of soils for particle-size analysis. Pochvovedeniye, 12, 55–68

Particle Size Analysis

61

Busacca AJ, Aniku JR and Singer MJ (1984) Dispersion of soils by an ultrasonic method that eliminates probe contact. Soil Sci. Soc. Am. J., 48, 1125– 1129 Edwards AP and Bremner JM (1967) Dispersion of soil particules by sonic vibrations. J. Soil Sci., 18, 1 Feller C, Burtin G and Herbillon A (1991) Utilisation des résines sodiques et des ultra-sons dans le fractionnement granulométrique de la matière organique des sols. Intérêt et limites. Science du sol, 29, 77–93 Gregorich EG, Kachandski RG and Voroney RP (1988) Ultrasonic dispersion of aggregates: distribution of organic matter in size fractions. Can. J. Soil Sci., 68, 395–403 Hinds AA and Lowe LE (1980) Dispersion and dissolution effects during ultrasonic dispersion of gleysolic soils in water and in electrolytes. Can. J. Soil Sci., 60, 329–335 Hinds AA and Lowe LE (1980) The use of an ultrasonic probe in soil dispersion and in the bulk isolation of organo-mineral complexes. Can. J. Soil Sci., 60, 389–392 Ilnicki P and Matelska U (1984) Ultrasound application for dispersion of soil samples for particle size analysis. Roezniki Gleboznaweze, 35, 15–24 Mikhail EH and Briner GP (1978) Routine particle size analysis of soils using sodium hypochlorite and ultrasonic dispersion. Aust. J. Soil Res., 16, 241–244 Minkin MB, Mulyar IA and Mulyar AI (1985) An ultrasonic method of analysing of water extracts from soils. Pochvovedeniye, 3, 136–140 Moen DE and Richardson JL (1984) Ultrasonic dispersion of soil aggregates stabilized by polyvinyl alcohol and T 403-glyoxal polymers. Soil Sci. Soc. Am. J., 48, 628–631 Morra MJ, Blank RR, Freeborn LL and Shafil B (1991) Size fractionation of soil organo-mineral complexes using ultrasonic dispersion. Soil Sci., 4, 294– 303 Schulze DG and Dixon JB (1979) High gradient enzymatic separation of iron oxydes and other magnetic minerals from soils clays. Soil Sci. Soc. Am. J., 43, 793–799

Pipette Method
Andreasen AHM and Andersen J (1930) Etude de l’influence de la dilution sur les résultats de l’analyse granulométrique par sédimentation. Kolloid Z., 50, 217 Bloom PR, Meter K and Crum JR (1985) Titration method for determination of clay-sized carbonates. Soil Sci. Soc. Am. J., 49, 1070–1073 Godse NG and Sannigrahi AK (1988) Comparative study on methods of particle-size analysis for vertisols. J. Indian Soc. Soil Sci., 36, 780–783

62

Mineralogical Analysis

Indorante SJ, Follmer LR, Hammer RD and Koenig PG (1990) Particle-size analysis by a modified pipette procedure. Soil Sci. Soc. Am. J., 54, 560– 563 Krumbein WC (1935) A time chart for mechanical analyses by the pipette method. J. Sediment. Petrol., 5, 93–95 Miller WP and Miller DM (1987) A micro-pipette method for soil mechanical analysis. Commun. Soil Sci. Plant Anal., 18, 1–15 Oreshkin NG (1979) Device for taking suspension samples for the particle-size analysis of soils. Soviet Soil Sci. (Pochvovedeniye), 4, 136–138 Richter M and Svartz H (1984) Analisis granulometrico de suelos en escala reducida. Ciencia del suelo, 2, 1–8 Shetron SG and Trettin CC (1984) Influence of mine tailing particle density on pipette procedures. Soil Sci. Soc. Am. J., 48, 418–420

Hydrometer Method
American Society for Testing and Materials (1972) Standard test methode for particle-size analysis of Soils – D 422-463. Annual Book of ASTM, 1985 Barthokur NN (1986) Clay fraction determinations with Beta-ray gauge. Commun. Soil Sci. Plant Anal., 17, 533–545 Fontes LEF (1982) A new cylinder for sedimentation of soil suspension in the determination of the clay fraction by the hydrometer method. Revista brasileira de Ciencia do Solo, 6, 152–154 Gee GW and Bauder JW (1979) Particle size analysis by hydrometer, a simplified method for routine textural analysis and a sensivity test of measurement parameters. Soil Sci. Soc. Am. J., 43, 1004–1007 Johnson JE, Bowles JA and Knuteson JA (1985) Comparison of pretreatments and dispersants on clay determination by the hydrometer method. Commun. Soil Sci. Plant Anal., 16, 1029–1037 Sur HS and Kvkal SS (1992) A modified hydrometer procedure for particle size analysis. Soil Sci., 153, 1–4

Instrumental Methods
Arustamyants YEI (1992) Instrumental methods for determining the particle-size composition of soils. Scr. Tech., 101–117 Barth, HG (1984) Modern Methods of Particle Size Analysis., Wiley, New York, 209 pages Cooper LR, Haverland RL, Hendricks DM and Knisel WG (1984) Microtrac particle-size analyzer: an alternative particle-size determination method for sediment and soil. Soil Sci., 132, 138–146

Particle Size Analysis

63

Devyatykh GG, Karpov YU A, Krylov VA and Lazukina OP (1987) Laser-ultra microscopic method of determining suspended particles in high-parity liquids. Talanta, 34, 133–139 Hendrix WP and Orr C (1970) Automate sedimentation size analysis instrument. Particle Size Analysis, 133–146 Hutton JT (1955) A method of particle size analysis of soils (balance de Plummet). CSIRO, Report, 11/55. Karsten JHM and Kotze WAG (1984) Soil particle analysis with the gamma alternation technique. Commun. Soil Sci. Plant Anal., 15, 731–739 Kirkland JJ and Yau WW (1983) Simultaneaous determination of particle size and density by sedimentation field flow fractionation (FFF). Anal. Chem., 55, 2165–2170 Kirkland JJ, Rementer SW and Yav WW (1981) Time-delayed exponential field-programmed sedimentation field flow fractionation for particlesize distribution analysis. Anal. Chem., 53, 1730–1736 Marshall TI (1956) A Plummett Balance for measuring the size distribution of soil particles. Aust. J. Appl. Sci., 7, 142–147 Mc Connel ML (1981) Particle size determination by quasielastic light scattering. Anal. Chem., 53, 1007–1018 Novich BE and Ring TA (1984) Colloid stability of clays using photron correlation spectroscopy. Clays Clay Miner., 32, 400–406 Oakley DM and Jennings BR (1982) Clay particle sizing by electrically induced birefringence. Clay Miner., 17, 313–325 Pennington KL and Lewis GC (1979) A comparison of electronic and pipet methods for mechanical analysis of soils. Soil Sci., 28, 280–284 Rybina VV (1979) Use of conductimetry for the determination of the particlesize composition of soils. Pochvovedeniye, 7, 134–138 Salbu B, Bjornstad HE, Linstrom NS, Lydersen E (1985) Size fractionation techniques in the determination of elements associated with particulate or colloidal material in natural fresh waters. Talanta, 32, 907–913 Svarovsky L and Allen T (1970) Performance of a new X-Ray sedimentometer. Particle Size Analysis, 147–157 Yang KC and Hogg R (1979) Estimation of particle size distributions from turbidimetric measurements. Anal. Chem., 51, 758–763 Yonker CR, Jones HK and Robertson DM (1987) Non aqueous sedimentation field flow fractionation. Anal. Chem., 59, 2574–2579

3

Fractionation of the Colloidal Systems
3.1 Introduction
The identification and the quantification of the finest soil fractions are essential to explain the transformation of minerals, as these fractions are directly related to pedogenesis and, in agronomy, to potential fertility. The nature and properties of these particles are of interest to agronomists (soil chemistry and physics: textural class, fertility, pore system, water storage, cohesion, slaking, etc.), soil scientists (pedogenesis, characterization and functioning of soils, lithological nature, products of alteration, etc.), geologists of the quartz period (sedimentology: origin of wind, marine, or lake deposits, typology of volcanic ash and heavy minerals, etc.), mineralogists and geochemists (assessments of alterations, mineral stocks liable to alteration, origin and nature of materials, etc.). The majority of the instrumental methods used for the determination of the texture of the soils (cf. Chap. 2) do not enable isolation of the fractions measured, but discontinuous methods based on sedimentation do enable reuse of the sand, silt and clay fractions, as long no contaminating dispersants are used. In practice, the limit of simple gravity methods is fractionation up to approximately 1 µm. Under certain conditions, ultracentrifugation makes it possible to reach the nanometric domain. Fractions below 0.5 µm contain practically no more quartz or primary minerals. Fractions below 0.2 µm enable better characterization of argillization horizons. This threshold, proposed around 1931, was at that time regarded as representative of the limit of “the colloidal state” because after elimination of oxides and hydroxides, the fraction presented homogeneous chemical composition comparable to a mono-dispersed system, i.e. the same exchange capacity, the same structural composition of the 0.2 µm, 0.1 µm, and 0.05 µm particles. When the final purpose of fractionation of the particles is physical or chemical determination, the different treatments that are carried out to put

66

Mineralogical Analysis

the very fine fractions in suspension can differ considerably from the methods used for textural analysis because secondary products cannot be significantly modified, in particular clays and oxides. In certain cases, it is possible to simply put fine fractions in suspension by ultrasound and to use ion exchange resin for desaturation. Dispersants of the hexametaphosphate or pyrophosphate type (cf. Chap. 2) should not be used. The criteria of Stokes law are suitable for centrifugation and the same problems will occur with particles whose speed is changing. By quantitatively isolating all the fractions, chemical and physical determination can be refined, and more detailed distribution curves established for the particle continuum.

3.2 Fractionation by Continuous Centrifugation

3.2.1 Principle After the treatments needed to isolate the primary particles (cf. Chap. 2), the sample is put in suspension for later analyses using a noncontaminating dispersant. The fine fractions (less than 2 µm) are first separated from the coarse fractions by several siphoning operations, which take the falling time of the silts into account. Five to six siphonings are sufficient for a quantitative recovery. Because of the volumes being dealt with, fractionation of the fine phases is carried out in a centrifuge with continuous inputs per ascensum made up of a vertical tube able to rotate at 52,000g (Fig. 3.1a). A transfer paper placed inside the tube makes it possible to collect the particles that sediment on the wall (Fig. 3.1b). The effluents are collected at the top of the bowl by a single or double chute (depending on the model). The suspension does not completely fill the tube but forms a concentric ring, the centre being a cylinder of air, R1 (Fig. 3.1b). The thickness of the film of suspension is R2 – R1. The particles gradually sediment on the walls of the bowl as a function of their density and of their diameter. They follow a parabolic trajectory starting from the point of deposit of the effluents. The radial application of the centrifugal force is accompanied by a vertical force. A particle of a given diameter and density will deposit according to the resulting force vector. Viscosity can be modified in the course of centrifugation by variations in temperature, the influence of the

Fractionation of the Colloidal Systems

67

pressures created at high speeds and possibly by the presence of thixotropic materials. With soils containing allophanes, it is sometimes possible to observe flocculation if the dispersing medium is not homogeneous throughout the iterative extractions, or due to loss of part of the swelling water. It should be noted that fibres of imogolite, mica plates, tests of diatoms, etc. do not follow Stokes’ law and are separated in a random way.

Fig. 3.1. (a) Uninterrupted ultracentrifuge system using compressed air – 45,000g (1 input nozzle for injection; 2, 3 exit chutes for effluents; 4 rev counter; 5 connecting with bowl; 6 centrifugation bowl). (b) Centrifugation bowl with transfer paper for recovery (division of the delivery points according to procedure in “Fractionation in Four Phases”).

The component perpendicular to the axis can be easily calculated starting from Stokes law (cf. Sect. 3.2.2), however, it should be noted that with centrifugation, the deposit rate of the particles is not constant due to variations in the intensity of the field of centrifugation which depend on the diameter of the rotor. The vertical component must be calculated as a function of the flow, which itself depends on the input, on the diameter of the injector channel, and on the diameter of the ring outfall.

68

Mineralogical Analysis

In practice, the formulas for computation of the standard Sharples centrifuge cannot be rigorously applied at the bottom and the top of the bowl because of turbulence at these levels (injector channel and deflector at the bottom, ring outfall at the top). Too much deposit can also modify the radius at the bottom of the bowl. However, the height of the deposit can be determined with sufficient accuracy by following a strict procedure. It may be advantageous to collect a relatively small quantity of sediments (not exceeding 10 g) at each treatment. Under certain operating conditions (number of revolutions, flow, etc.), the finest particles can pass across the bowl without sedimentation because the time of passage is insufficient for the centrifugal force to transfer them to the wall. However, if one of the variables is modified, for example, the flow is slowed down by changing the input tube; these particles can also be collected. With a low flow, separation is differentiated more satisfactorily than with a high flow. For satisfactory separation, it is better to use tubes with a small diameter and a not too high charge. For simple separation (for purification by enrichment) a 2 mm tube and a higher charge (75 cm for example) can be used. The drain may still contain fine particles depending on the number of revolutions applied. The system makes it possible to use the drain again with a lower flow rate and to collect the solid fraction on a new transfer paper placed in the bowl.
Table 3.1. Deposit of particles of various minerals by continuous ultracentrifugation –1 (flow 730 mL min speed 20,000g, viscosity 1) mineral density diameter of particles which sediment at a height of 10 cm in the bowl (in µm)
360
320 280 115

opal
gibbsite quartz hematite

2.10
2.40 2.65 (average used for the soils) 5.26

It is important to choose a speed and a flow which allow non-uniform deposit to obtain satisfactory classification of the particles. It is difficult to include all the variables in the calculation, and the forecast will be uncertain if the average density of 2.65 used for the soils is not regular. Indeed, if a mixture of minerals of different densities is used whose

Fractionation of the Colloidal Systems

69

micro-particles are individualized in the suspension, very different particles with the same diameter will be found at any level (Table 3.1). At one time, it seemed that graphic methods would be both adequate and fast. Saunders (1948) studied a nomographical representation applicable to continuous ultracentrifugation (1936–1940), and was able to determine that certain factors in the equation of Hauser and Lynn (cf. Sect. 3.2.2) remained constant for a given procedure thus allowing simplification and a move to a nomographical representation with five variables that could be extended to six variables using the method of Davis (1943). It thus became possible to rapidly determine the height of deposit of elementary particles with different densities but the same diameter.

3.2.2 Theory The method of Hauser and Lynn (1940) to calculate the size of the particles is one of the most powerful for use with an ultracentrifuge with continuous input. The equation makes it possible to express the vertical distance (Y in cm) from the deposit of a particle of a given size, measured starting from the bottom of the bowl of the centrifugal machine by

Y =C where

18 K 1Q η π (R − R12 )D 2ω 2δρ
2 2

(3.1)

(3.1’) 2 2 4 X0 X0 R2 = distance from the axis of rotation to the side of the bowl (cm); R1 = distance from the axis of rotation on the surface of the liquid in the bowl (cm); X0 = distance from the axis of rotation at which a given particle must start to deposit on the side of the bowl (cm); K1 = function of the construction of the bowl (cm–2) 2 R 2 − R12 = (3.2) 2 (3 / 4)R14 + (1 / 4)R 24 − R12 R 2 − R14 ln(R1 / R 2 ) Q = throughput speed (flow of the suspension mL s–1); η = viscosity of the dispersion medium (poise); D = sphere equivalent diameter of the particles which sediment at Y cm; ω = angular speed of rotation (radian per second);

C =

2 R2

ln

R2

R12

ln

R2

2

+

X 02 − R22

70

Mineralogical Analysis

δρ = difference in density between the dispersed particles (Table 3.2) and the medium of dispersion (g mL–1). Under standard conditions (flow and range of the particles), the equation of Hauser and Lynn is related to X0 and D. For particles of a given diameter, the equation is a function of only X0 , which makes it possible to plot the curve of C vs Y. On this basis, Saunders (1948) established a system of monograms which allows the equivalent diameter of the particles settling in the bowl to be calculated with satisfactory precision. 18 K 1 By defining the constant A = , (3.1) becomes: 2 R 2 − R12

(

)

Y A η Q = 2 2 C D ω δρ

(3.3)

Table 3.2. Specific density of some minerals (average density used for the soils = 2.65) minerals density minerals density

Al

boehmite ALO(OH)

3.07

Fe

akaganeite 3+ Fe O(OH, Cl)

3.55

diaspore AlO(OH) bayerite Al(OH)3 gibbsite Al(OH)3 nordstrandite Al(OH)3 corundum Al2O3 akdalaite 4Al2O3, H2O bauxite Al2O3, 2H2O Mn manganosite MnO

3.40 2.53 2.40 2.43 4.10 3.68 2.55 5.36 clays

goethite FeO(OH) lepidocrocite FeO(OH) hematite Fe2O3 maghemite Fe2O3 magnetite Fe3O4 (Fe2+Fe23+O4) ilmenite FeO–TiO2 pyrite FeS2 kaolinite

4.0– 4.4 3.85 5.26 5.10 5.17 4.44– 4.90 5.02 2.60

pyrolusite MnO2 ramsdelite MnO2 manganite MnO(OH) feitknechtite MnO(OH) groutite MnO(OH)

5.06 4.50 4.33 3.80 4.15

halloysite dickite montmorillonite nontronite beidellite

2.10 2.60 2–3.00 2–3.00 2–3.00

Fractionation of the Colloidal Systems
pyrochlorite Mn(OH)2 nsutite Mn+2Mn+4O2(OH)2 hausmanite Mn+2Mn2+3O4 3.25 4.50 4.84 micas: biotite phlogopite glauconite muscovite 3.00 2.80 2.60 2.80 2.6– 2.9

71

illite Si coesite SiO2 2.93 feldspar white feldspar Na2O, Al2O3, 6SiO2
andesine (CaO, Na2O Al2O3, 4SiO2 oligoclase NaAlSi3O8 + CaAl2Si2O8 orthoclase KAlSi3O8

2.61– 2.64
2.65– 2.69 2.62– 2.67 2.56

cristobalite SiO2 quartz SiO2 tridymite SiO2 silhydrite 3SiO2, H2O opal SiO2, nH2O Ca calcite CaCO3
aragonite CaCO3

2.33 2.65 2.26 2.14 2.10 2.7–2.9
2.9

gypsum CaSO4, 2H2O

2.3

This equation can be treated using the “nomographical method” according to Davis (1943) and gives a value of Y/C to each value of Y. If η is expressed in centipoises, Q in mL min –1, D in nm, ω in rotations min–1, δρ in g mL–1 Y in cm and C in cm2 the equation is written: Y Aη Q = 1.52 × 1012 2 2 C D ω δρ (3.3’)

This equation can be reduced to four equations with three variables and three parameters α, β and γ : log α = log η − log Y C log β = log δρ − 2 log ω log γ = log α + log β 2 log D = log γ + log Q The nomographical method of Davis (1943) provides a solution by tracing four lines which represent the solution of one of these equations (Fig. 3.2). Constant A is included starting from a scale representing a numerical solution of (3.3’). For example, for a centrifugation tube where R1 = 2.175 cm and R2 = 0.735 cm, constant A is equal to 2.44 × 1012. The final formula for direct calculation will be:

72 Mineralogical Analysis

Fig. 3.2. Nomographer: (1) trace 1–2 while cutting A, (2) trace 5–6 while cutting C, (3) trace AC while cutting B, (4) trace B-3 while cutting 4, the equivalent diameter is obtained in nanometres of the particles deposited at the height selected (Y) in the centrifugation bowl

Fractionation of the Colloidal Systems

73

D = 2.44 ×10 12

(ω δρY C )
2

ηQ

1

2

with value Y/C for the height Y and the units of (3.3’).
3.2.3 Equipment and reagents Equipment

– Standard Sharples T1 ultracentrifuge with continuous circulation and a turbine equipped with an 8RY stainless steel bowl with a diameter of 44 mm – 6 L Pyrex bottle with broad neck with stopper – Stopper pierced with a glass tube with an interior diameter of 10 mm cut in bevel (for input of suspension) – 12 cm Conical forceps with jaw punts (to retrieve the transfer paper) – Small plastic spatula – Stylet (to lift the transfer paper for removal with the grip)

Fig. 3.3. Spectra of X-ray diffraction: dashed line = Tymek support alone, solid line = kaolinite on Tymek support (sufficiently thick to avoid contamination).

74

Mineralogical Analysis

– Plastic transfer papers with frosted interior (Integraph, Invar 75 µm, Kodatrace, Chronaflex or Tymek Dupont de Nemours) 204×150 mm sheets; the thickness of this support must be homogeneous, it must have a constant weight per surface unit, be resistant to water, have a flat DRX spectrum or well-defined peaks outside the zones of measurement of the sample (Fig. 3.3) – 1/10 mg balances – Set of suitable pillboxes and micro-bottles (plastic, single use, Fig. 3.4)

Fig. 3.4. Example of system for processing the fine fractions separated on transfer paper (Fig. 3.1), 1 = regrouping on MEB support, 2 = only andisols– histosols, 3 = pillbox, micro-bottles

– Aluminium supports for scanning electron microscopy
Reagents

Cf. Chap. 2

Fractionation of the Colloidal Systems

75

3.2.4 Procedure

Standard Continuous Ultracentrifugation

– Choose the diameter of the injector channel, the diameter of the ring outfall, and the height h of level X (Fig. 3.1). – Place in the bowl of the centrifuge a tared transfer paper with the selected cutting plan drawn on the back (the weight of the transfer paper P0 makes it possible to calculate the weights P01, P02, P03, corresponding to the respective surfaces of zones A, B, C in Fig. 3.1). – Suspend the bowl on the rotor and place the container for the recovery of the effluents under the chute. – Fill the funnel to a constant level X with the same dispersant as the samples, taking care not to trap air in the adduction tube. – Homogenize the bottle containing the clay suspension. – Remove an aliquot of 2 mL (for grid TEM). – Rock the bottle on the funnel; switch on the centrifugal machine at the selected speed. – Place the injector channel under the bowl and open the input cock. When all the liquid has gone, add 200 mL of the dispersion liquid to drive out all the suspension remaining in the bowl (approximately 150 mL). Adjust to a maximum speed of 52,000g for 2–3 min to stabilize the deposit. Disconnect the input tube and stop the centrifuge (collect the liquid remaining in the tube in a crystallizer and discard it if it is clear). Remove the transfer paper carefully by holding the bowl obliquely to not contaminate the top of the bowl with coarse particles. Spread the paper out flat. Recover any trace of deposit on the deflector and add it to the bottom of the transfer paper. Leave to dry at room temperature (if necessary recover wet clay with a spatula and place it on the appropriate zone of the transfer paper before drying). Weigh the transfer paper and dried clay: P1. Deposited clay corresponds to P1 – P0 Cut the transfer paper following the plan on the back. This makes it possible to weigh zones A, B, C (Fig. 3.1), i.e. P11, P12, P13.
Continuous Fractionation of the Colloidal Particles

The complex and time-consuming method of Hauser and Lynn (1940) enables isolation of the fine fractions from a suspension (Fig. 3.5) and the

76

Mineralogical Analysis

establishment of cumulative curves of distribution of the particles which can nowadays be accomplished continuously with an automatic apparatus for the measurement of particle size (cf. Chap. 2). This type of apparatus has two centrifugation speeds and seven different flows, i.e. 11 passages in the centrifuge to classify particles between 1 µm and 24 nm with a Sharples continuous centrifuge with a turbine. This method cannot be used for routine tasks because of the length of the operations, or for fragile samples that are difficult to maintain in suspension like soils with allophane. The method is nevertheless useful in metallogeny to identify the enriched fractions (release mesh). The suspensions must be diluted to accomplish fractionation without an awkward piston effect.
Fractionation in Four Phases

The method of Gautheyrou and Gautheyrou (1967) is based on the equations of Hauser and Lynn and the nomographical system of Saunders (cf. Sect. 3.2.2). The transfer paper shown in Fig. 3.1b is an example of one layout suited to the needs of mineralogical analysis, but other alternatives are possible considering that the particles deposited in a horizontal plane are similar and that the solid phase varies upwards. Zones A, B, C (Figs. 3.1 and 3.4) allow separation of the particle size phases whose significance depends on the procedure used (these fractions are used for chemical and physical analyses). Zones A’ and C’ (Figs. 3.1 and 3.4) enable fractions to be isolated for X-ray diffraction either after crushing for powder diagrams, or pretreatments, or directly for oriented diagrams on the 24 × 24 mm2 (cf. Chap. 4). After sticking the 5 × 5 mm2 on a suitable support with carbon lacquer (cf. Chap. 8) they can be used for electronic scan microscopy with EDX microanalysis. After dilution and preparation of the grids, a 1 mL sample of each suspension enables observation by transmission electronic microscopy (cf. Chap. 8).
Adjustments Charge h: 75 cm (Fig. 3.1); temperature: 20°C; tube: 2 mm diameter; flow: 730 mL min–1; time of passage in the bowl: 15 s; average density: 2.65; ring outfall: 0. Charge, flow, temperature remain constant; only the speed is modified at each centrifugation: 6,000, 10,000, 25,000, 50,000g (at 50,000g, if the sample contains very fine particles, it may be necessary to use a 1 mm tube corresponding to a flow of 160 mL per minute to fix all the very fine phase).

21,000g

51,000g

1,200

1,200

Fractionation of the Colloidal Systems

Fig. 3.6. Diagram of fractionation by method “Fractionation with Different Flows”

decantation

soil suspension
-1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1

Fig. 3.5. Diagram of fractionation by method “Continuous Fractionation of the Colloidal Particles”

77

78

Mineralogical Analysis

Fraction 2–1 µm This fraction is separated at a centrifugation speed 6,000g and recovered on the first transfer paper (TP): weight TP alone: P01, TP + deposit: P11, weight clay: 2–1 µm: P11–P01 : Pc1 Express in % compared to the initial weight of soil. Fraction 1–0.5 µm Recovered on the second TP at 10,000g: Pc2 Fraction 0.5–0.2 µm On the third TP at 25,000g: Pc3 Fraction 0.2–0.05 µm On the fourth TP at 50,000g and lower flow of 160 mL min–1, 1 mm tube: Pc4 Clay < 0.05 µm This very fine clay is contained in the last draining water and can be recovered by flocculation. This fraction is generally a clear extract that can be discarded because it contains the majority of the residual impurities of the reagents used for the initial preparation of clay. Electronic microscopy can also be used, but concentration by evaporation can result in risks of hydrolysis and neo-formation. Remarks If the deposit is too thick, irregular cracking can occur in clays with a high shrinkage coefficient, which makes it impossible to cut the 5 × 5 mm2 for oriented XRD. In this case, the squares are not used but instead a sliver of clay is stuck down with carbon lacquer. In certain soil samples the finest fractions are practically non-existent. In this case it is possible to stop at the second or the third centrifugation. If major variations are observed between the coarsest phase and the finest phase, preparation and analyses of sub-samples should be performed progressively. Certain studies may require conservation of part of fresh clay without drying. In this case, the transfer paper should have an additional vertical separation. Fresh clay should be recovered immediately at the outlet of the bowl with a plastic spatula and quickly stored in a pillbox. Quantification can be checked on the fraction which is dried but there is a risk of error due to transformations during drying.

Fractionation of the Colloidal Systems

79

The same method can also be used to: – Separate and enrich two phases; for example two phases with the same mineral density but different particle size (micro-micas and coarser crystallized kaolinite). – Separate two products of the same particle size but different density, for example aluminous products with a density <3.0 and ferrous or ferric products with a density of 4–5. – Collect the particles of a given diameter and density on a narrow zone (metallogeny – separation meshes), etc. Chlorite, vermiculite, or kaolinite enrichment is often observed in the coarse fractions. Well-crystallized kaolinite is generally present in fractions >0.5 µm. Quartz and muscovite are almost eliminated in the fractions <0.5 µm, which makes it possible “to clean” the spectra. Gibbsite is mostly retained in the coarse fractions and is only present in very small amounts in the very fine phases. Halloysite enrichment can be also observed, as well as enrichments in substances with short distance crystalline arrangement in fractions <0.5 µm or 0.2 µm, whereas well-crystallized kaolinite disappears. Deferrization increases the smoothness of the particles revealing the incorporating effect of iron. Studying the different fractions by X-ray diffraction enables observation of possible crystallochemical heterogeneity of the colloidal fraction and identification of mineral filiations which occurred during evolution and weathering. Minerals whose spectral signature is readable only above 5% (for example Sepiolite) can also be detected.
Fractionation with Different Flows

In this technique (Biological Centre of Pedology, Nancy, France), the clay suspension >2 µm at a concentration not exceeding 1%, is subjected to an initial series of centrifugation at fixed speed and flows (Fig. 3.6). A second series of centrifugation with a much lower flow (31 mL min–1) makes it possible to separate the finer fractions. Thanks to repeated centrifugation and the weak concentration of the medium, the separation of the fractions is considered quantitative.

3.3. Pretreatment of the extracted phases

Mineralogical analyses performed directly on total soil samples provide information on all the most abundant components, but do not allow

80

Mineralogical Analysis

detection of the presence of low concentration phases because of the lack of sensitivity of the instrumental techniques and the occurrence of much interference. These low-concentration phases (<5 %), which can be highly significant in explaining certain processes of soil genesis, are masked by background noise even when they are above their threshold of detection. It is thus necessary to eliminate interference and to concentrate the “mineralogical clay” phase. Using a suitable fractionation method, clays are purified and concentrated and appear in a homoionic form: NH4+, Ca2+, or H+ depending on the case. As mineral cements and the organomineral links have been destroyed, it is possible to obtain satisfactory separation of the particles. The use of ultrasound enables good separation of elementary particles, and complementary pretreatments can be performed on clays to allow the use of specific instrumental techniques. Selective dissolution makes it possible to eliminate the iron oxides which can obstruct XRD (cf. Chap. 4), fluorescence with a copper tube) and DTA–TGA (cf. Chap. 7), oxidation Fe2+). Different pretreatments make it possible to carry out analyses using NMR, ESR, Mossbauer, etc. Gels and substances that are amorphous to X-ray and have crystal lattices with short distance arrangement can be dissolved, enabling their quantification and the production of differential spectra (DXRD) to identify them. Any calcium carbonate that is still present can be eliminated by complexing Ca2+ ions with a solution of normal EDTA. Pretreatments also help improve the orientation of clays (which is disturbed by iron, for example in coatings), the intensity of the spectra of diffraction and the ratio of diffraction to background noise. Clays often have to be studied after homoionic saturation by a cation (such as Mg2+ which regulates the adsorption of water by clays with an expansible interfoliaceous space, or K+ which limits the adsorption of water and thus the swelling of the layers). Other treatments, like solvation by polar solvents, or the creation of complexes of intercalation, make it possible to identify certain clays. Heat treatments are also used specifically to cause the collapse of the lattices or to modify surface properties. These methods are described in detail in Chaps. 4–7.

Fractionation of the Colloidal Systems

81

References
Davis DS (1943) Empirical Equations and Nomography. Mc Graw Hill, New York, 1, 104–114 Gautheyrou J and Gautheyrou M (1967) Mode opératoire pour l’extraction et la purification de la fraction argileuse < 2 µm. Notes de laboratoire, Orstom-Guadeloupe, mars 1968, 1–9, Orstom Hauser EA and Lynn JE (1940) Separation and fractionation of colloidal systems. Ind. Eng. Chem., 32, 659–662 Saunders E (1948) Nomograph for particle size determination with the Sharples supercentrifuge. Anal. Chem., 20, 379–381

Bibliography
Atterberg A (1912) Die mechanische bodenanalyse und die klassification der mineralböden schwedens. Intern. Mitt. Bodenk, 2, 312–342 Coca Prados J and Bueno de las Heras J (1977) Dinamica de particulas en suspensions solido–liquido. I – Sedimentacion de particulas. Ingeniera quimica, 153–162 Colmet-Daage F, Gautheyrou J, Gautheyrou M, Kimpe de C, Fusil G and Sieffermann G (1972) Dispersion et étude des fractions fines de sols à allophane des Antilles et d’Amérique latine. IIème partie : Modifications de la nature et de la composition de la fraction inférieure à 2 microns selon la taille des particules. Cahiers Orstom, série. Pédol., X, 219–241 Davis JM (1986) General retention theory for sedimentation Field-FlowFractionation. Anal. Chem., 58, 161–164 Essigton ME, Mattigod SV and Ervin JO (1985) Particles sedimentation rates in the linear density gradient. Soil Sci. Soc. Am. J., 49, 767–771 Gautheyrou J and Gautheyrou M (1982) Fractionnement des systèmes colloïdaux argileux par centrifugation continue. Notes laboratoire Orstom Bondy, 1– 38 Hauser EA and Reed CE (1936) Studies in thixotropy. I – Development of a new method for measuring particle-size distribution in colloidal systems. J. Phys. Chem., 40, 1169–1182 Horrocks M (2005) A combined procedure for recovering phytoliths and starch residues from soils, sedimentary deposits and similar materials. J. Archaeological Sci., 32, 1169–1175 Jackson ML, Whittig LD and Pennington RP (1949) Segregation procedure for the mineralogical analysis of soils. Soil Sci. Soc. Am. Proc., 14, 77–81 Jacobsen AE and Sullivan WF (1946) Centrifugal sedimentation method for particle size distribution. Ind. Eng. Chem., 18, 360–364 Jaymes WF and Bigham JM (1986) Concentration of iron oxides from soil clays by density gradient centrifugation. Soil Sci. Soc. Am. J., 50, 1633–1639

82

Mineralogical Analysis

Johnson L (1956) Particle size analysis and centrifugal sedimentation. Trans. Bull. Ceram Soc., 55, 267–285 Kamack HJ (1951) Particle size determination by centrifugal pipet sedimentation. Anal. Chem., 23, 844–850 Kittrick JA and Hure EW (1963) A procedure for the particle-size separation of soils for X-Ray diffraction analysis. Soil Sci., 96, 5, 319–325 Koch T and Giddings JC (1986) High-speed separation of large (> 1 µm) particles by steric Field-Flow-Fractionation. Anal. Chem., 58, 994–997 Levitz PE (2005) Confined dynamics, forms and transitions in colloidal systems: from clay to DNA. Magn. Reson. Imaging, 23, 147–152 Marshal CE (1931) Studies in the degree of dispersion of the clays, I – Notes on the technique and accuracy of mechanical analysis using the centrifuge. J. Soc. Chem. Ind., SDT, 444–450 Muog E, Taylor JR, Pearson RW, Weeks AE and Simonson RW (1936) Procedure for special type of mechanical and mineralogical soil analysis. Soil Sci. Soc. Am. Proc., 101–112 Rouiller J, Brethes A, Burtin G and Guillet B (1984) Fractionnement des argiles par ultra-centrifugation en continu : évolution des illites en milieu podzolique. Sci. Géol. Bull., 37, 319–331 Schachman HK (1948) Determination of sedimentation constants in the Sharples supercentrifuge. J. Phys. Colloid. Chem., 52, 1034–1045 Tan KH (1996) Soil Sampling, Preparation and Analysis. Dekker, New York, 278–361 Tanner CB and Jackson ML (1947) Nomographs of sedimentation times for soil particles under gravity or centrifugal acceleration. Soil Sci. Soc. Am. Proc., 12, 60–65 Tran-Vinh-Ann and Ndejuru E (1972) Analyse granulométrique de la fraction argileuse par centrifugation en flux continu. Mise au point d’une méthode et application à quelques sols tropicaux. Pédologie, XXII, 366–382 Truo E, Taylor JR, Simonson RW and Week ME (1936) Mechanical and mineralogical subdivision of the clay separate of soils. Soil Sci. Soc. Proc., 175–179 Tu Y, O’Carroll JB, Kotlyar LS, Sparks BD, Ng S, Chung KH and Cuddy G (2005) Recovery of bitumen from oilsands: gelation of ultra-fine clay in the primary separation vessel. Fuel, 84, 653–660 Whittig LD and Allardice WR (1986) X-Ray diffraction techniques – separation of particle – size fraction. In Klute A (ed.), Method of Soil Analysis Part Physical and Mineralogical Methods, second edition, American Society or Agronomy, 340–342

4

Mineralogical Characterization by X-Ray Diffractometry

4.1 Introduction

4.1.1 X-Ray Diffraction and Mineralogy Methods using optical microscopy in petrography are not suitable for the identification of mineralogical clays with small particles whose crystal lattices vary with water content and with their ionic environment and whose chemical composition is often unclear (Tables 4.1 and 4.2). Among other available methods, X-ray diffraction (XRD) is one of the most efficient. Coherent scattering of the incidental radiation in XRD makes it possible to clearly identify both the parameters of the crystal lattice and the geometrical distribution of the atoms in the crystal mesh. XRD can be combined with or supplemented by geochemical and isotopic analyses (AAS, ICP, ICP-MS, EXAFS, etc.), thermal analyses (DTA-TGA, DSC, EGD, EGA, etc.), analyses that enable evaluation of interatomic or intermolecular binding energies and order–disorder relations (e.g. FTIR, Raman spectrometry, Mossbauer spectrometry, NMR), high resolution transmission electronic microscopy (+ electron microdiffraction) and electronic scan microscopy. EDX or WDX probes make it possible to link in situ chemical composition with the shapes of the particles to be observed. Total chemical composition is determined by total analyses after mineralization in mediums that enable solubilization of all the components; selective dissolution makes it possible to subdivide the sample into fractions of different chemical resistance; these sub-divisions are essential both for quantification and for purification of the samples before analysis by instrumental methods (e.g. XRD, IR or NMR spectroscopy).

84

Mineralogical Analysis

Table 4.1. Classification of clays proposed by the international association for the study of clays (AIPEA1)

type

group (x = charge sub-group (n = number of by unit formula) cations of octahedral layers)

species

1:1

kaolinite – serpentine (x = 0) pyrophyllite – Talc (x = 0) smectites montmorillonite saponite (x = 0.25–0.6)

kaolinite (n = 2) serpentine (n = 3)

kaolinite, halloysite, chrysotile, lizardite,
antigorite

pyrophillite (n = 2) talc (n = 3) dioctahedral smectites or
montmorillonites (n = 3)

pyrophillite
talc

montmorillonite, beidellite, nontronite saponite, hectorite, sauconite

trioctahedral smectites or saponite (n = 3)

2:1

vermiculite (x = 0.6–0.9)

dioctahedral vermiculites (n = dioctahedral 2) vermiculite

trioctahedral vermiculites (n = trioctahedral 3) micas2 (x = 1) breakable micas (x = 2) dioctahedral micas (n = 2) vermiculite muscovite, paragonite

trioctahedral micas (n = 3) biotite, phlogopite dioctahedral breakable micas margarite (n = 2) trioctahedral breakable micas clintonite (n = 3)

chlorite (x variable) 2:1:1

dioctahedral chlorites (4<n<5) donbassite di-trioctahedral chlorites cookeite, sudoïte

trioctahedral chlorites (5 < n < 6) penninite, clinochlore, prochlorite

Inter-analytical tests can be performed at different levels: on the structure of clays (geochemical relations, pedological differentiation
1

AIPEA Association Internationale pour l’Etude des Argiles, GPO Box 2434, Brisbane, Qld 4001 Australia 2 Illites (or hydromica), Sericite, etc., many materials labelled illites can be interstratified.

X-ray Diffractometry

85

within a profile and spatial differentiation), hydrous properties (porosity, permeability, functional waterlogging generated by the nature and the proportion of clays), adsorbing complexes (charge distribution, CEC, etc.). For a detailed quantitative study, in addition to the clay particle fraction, it is generally necessary to analyse the fine silt fraction which can contain interstratified minerals, particularly in the case of micas and well crystallized kaolinite. A balance takes into account clay and s associated phases, oxides, hydroxides, etc. making it possible to explain apparently unmatched results (excessive Si4+ content due to diatoms, very fine quartz, high percentages of K+ originating from potassic feldspars, micas, etc.).
Table. 4.2. Structural lexicon (AIPEA, 1972) English (atomic) plane (tetrahedral or octahedral) sheet (plane combination) 1:1 or 2:1 layer (sheet combination) French plan (atomique) couche (tétraédrique ou octaédrique,combinaison de plans) feuillet 1:1 ou 2:1 (combinaison de couches) interlayer space espace interfoliaire unit structure = combination of layers assemblage de feuillets + matériel inter+ interlayer materials foliaire = unité Structurale lattice Réseau

Fig. 4.1. The crystal mesh

86

Mineralogical Analysis

4.1.2 Principle A crystal is defined as a solid made up of atoms assembled in a threedimensional periodic model. Lengths a, b, c and angles α, β, γ between the planes define the mesh parameters of the basic unit (Fig. 4.1). When monochromatic X-ray beams of suitable wavelength strike a crystalline plane, the X-rays are reflected by the atoms of the crystal. The signal is reinforced in a particular direction if the rays reflected by the different planes (Fig. 4.2) are in phase. This phenomenon corresponds to Bragg’s law

2 d sin ϑ = nλ

(4.1)

where d is the space between atomic planes or the inter-reticular distance in the crystal (d(hkl)); λ is wavelength and θ is angle between beam and atomic plane and n is the order of diffraction (integer number). All the planes of a crystal diffract the X-ray when the crystal is tilted at certain angles θ of the incidental beam of wavelength λ in accordance with the law (4.1). The angles θ are linked to wavelength λ and distance d, which are expressed in Angstroms or nanometres (1 Å = 0.1 nm = 10 –10 m). If the wavelength is known, measuring the angle of reflection makes it possible to determine the inter-reticular spaces. Remarks Certain minerals can be “amorphous” to X-ray either because they do not have a specific crystalline arrangement (true of glasses) or because they have short-range organization that is too small to be detected at a wavelength of 1–2 Å. XRD is not the best technique for the study of non-crystalline solids such as allophanes which are made up of clusters of Si atoms presenting structural elements with interlayer distances corresponding to 1 or 2 neighbouring atoms. Atoms of silicon in a tetrahedral position and atoms of aluminium in octahedral coordination but with no regular symmetry, cannot give well defined peaks, but only broad and badly defined peaks that appear around 0.33 and 0.22 nm.

X-ray Diffractometry

87

4.1.3 XRD Instrumentation X-rays were discovered in 1895 by Roentgen, but the phenomenon of crystal diffraction was discovered only in 1912.

Fig. 4.2 Diffraction of an incidental beam by crystalline reticular planes.Lines p, p1, p2 represent the parallel and equidistant reticular planes separated by space d. An X-ray beam striking the higher plane p will be reflected in the incidental angle θ. To obtain a measurable reflection, all the rays reflected by planes p, p1, p2, etc, must be in phase. To achieve this, GE + EH, the path difference between radiations ABC and DEF must be equal to a whole number of the wavelength. As GE = EH = d sin θ, the condition is thus the law of Bragg (1).

X-rays are located between UV and gamma radiation wavelengths, i.e. approximately 10 to 10– 2 nm. In XRD, “hard” radiation is used with wavelengths from 1.5 to 1.9 Å depending on the anti-cathode used (1 Å = 0.1 nm). Radiation X-ray is propagated in a straight line. It is necessary to ensure the radiation is as monochromatic as possible by using a set of filters, slits and a monochromator. The apparatus comprises: – a generator with stabilized high voltage and micro-intensity which supplies the X-ray tube and the counter – a sealed X-ray tube, including a source of electrons maintained from 20 to 50 kV by means of a high negative potential and an anode or anticathode (Table 4.3) made of thick metal, cooled by a water circuit;

88

Mineralogical Analysis

high speed bombardment of the electrons causes transfer of energy to the atoms of the anode-target bringing them to a higher energy level, thus creating orbital vacancies of electrons; the quantum of energy produced is characteristic of the atoms of the anode; X-ray photons leave the tube by 300 µm thick beryllium windows that are transparent to the X-ray (the spot must be as small as possible to concentrate the energy of the electrons on a limited zone of the anode and to ensure a high intensity X-ray source); the power of the tube is limited by the quantity of heat likely to be dissipated by the anode and is expressed by the maximum acceptable value in mA for a given voltage (1–3 kV or more in the case of a rotating anode); characteristic radiations are obtained only starting from a given critical voltage of excitation, thus strict regulation of the voltage and intensity is required to avoid modifying the wavelength – a goniometer, which makes it possible to rotate the sample and the counter under the conditions fixed by the Bragg equation; this provides a support both for the sample whose plane is adjusted very precisely and for the focused detector which turns around the same axis in the same direction at a suitable speed ratio – a linear beam is obtained by means of the Soller slits and the degree of divergence which limits the opening of the beam; variable slits are now used with openings linked to the angle making it possible to irradiate a constant surface, which is particularly useful with small angles because the effects of the direct beam are limited – a reception slit limits the width of the beam in the focal plane; the narrower the slit, the higher the resolution, though there is a loss in intensity; a graphite back monochromator limits fluorescence radiation, incoherent radiations and the Compton effect; the perfect alignment of all the different elements determines the quality of the measurements – a detection system (counter) makes it possible to measure the intensity of the X-ray transmitted; the number of pulsations per unit of time is proportional to the quantity of X-ray transmitted; the counter can be linear or proportional, or detection can be by scintillation, or occasionally by semiconductor (semiconductor counter must be kept in liquid nitrogen at −196°C). Safety X-ray radiation is dangerous and can cause burns, genetic modifications, cancers, etc. The risks associated with high voltage must be also taken into account. Careful prevention is essential and strict regulations apply to all apparatuses, which must be equipped with safety devices:

X-ray Diffractometry

89

– the sealed tube must be protected by thick walls to eliminate risk of radiation – the goniometer must be insulated with lead glass or lead plastic to protect the whole apparatus; it should only be opened when the tube is switched off – the operator must wear a monitoring badge with a sensitive film (dosimeter) that accounts for possible whole-body irradiation (which must be checked regularly), as well as a ring to measure irradiation of the hands – operators must have regular medical check-ups to detect changes in the blood count (white blood cell count, etc.) – a Geiger counter must be used to check for radiation leaks: fluorescent screens should be placed on supports made of zinc doped with nickel to identify the zones struck by the beam while alignment is being adjusted.
Table 4.3. Characteristics of some anti-cathodes
α antiKβ wavelength cathode filter

K

Å

excitation operating induced potential potential remarks fluorescence (KV) (KV)

Cu

1.542

Ni

Co–Fe

9.0

25–45

penetration and average dispersion, not much affected by air

Co

1.791

Fe

Mn–Cr

7.8

25–35

weak penetration capacity, great dispersion, not much affected by air Weak penetration capacity, great dispersion

Fe

1.947

Mn

Cr–V

7.1

25–35

90

Mineralogical Analysis

4.2. Qualitative diffractometry

4.2.1 Overview of Preparation of the Samples

See Fig. 4.3.

-

Fig. 4.3. Preparation of samples for X-ray diffractometry

4.2.2. Preparation for Powder Diagrams

Objective

This method is a general way to identify mineral species without preferential orientations. It enables quantitative analysis as the

X-ray Diffractometry

91

relative intensity of maxima diffractions is approximately proportional to the number of crystals present if the level of anisotropy is low. Principle In the method based on the “spectrum of random powder” (Thomson et al., 1972; Peterson et al., 1986; Decarreau, 1990), the crystal is examined in the form of a fine isotropic powder under a monochromatic X-ray beam. Random orientation must statistically represent all possible orientations of the different particles and provide a complete spectrum of minerals likely to diffract the X-ray (clays, oxides, hydroxides, oxyhydroxides, non-weathered primary minerals, various salts, etc.). Each particle is considered to be a micro-crystal or an assembly of micro-crystals and the powder mass can be compared with a single crystal turning not in only one axis but in all possible axes. The powder spectrum makes it possible to fix the relative intensities of the peaks 3 indexed by JCPDS and for this reason it should always be carried out before any other operation. After compression in the support at the semi-microscale, the “powder” sample includes micro-aggregates of fractal dimensions with a piled-up structure with open porosity – the grains being only very slightly connected – and its characteristics depend on the crushing, compression and homogeneity of the medium. The total average orientation at this scale is very weak. On the other hand, at the sub-microscale (a few tens or hundreds of nanometers), the primary morphological units are mixed stacks of clayey crystallites and can consequently present an orientation with a varying degree of disorder depending on the types of clay present. The use of a revolving support improves the level of randomization, but requires a relatively large quantity of powder, i.e. approximately 400 mg. Procedure A powder diagram can be performed on whole soil or on soil fractions (cf. Chap. 3). Clay obtained by centrifugation (centrifugation pellet or plastic transfer paper) is air dried and then crushed in an agate mortar to obtain a homogeneous powder: – With a spatula place in a hollow support (Fig. 4.4) the quantity of powder needed to almost fill the cavity.
3JCPDS – ICDD = Joint Commitee on Powder Diffraction Standards – International Center for Diffraction Data, Newtown Square Corporate Campus, 12 Campus Bdv, Newtown Square Pennsylvania 19073-3273 (USA).

92

Mineralogical Analysis

– With a ground glass slide, gently flatten the powder – Gradually add powder to fill the remaining cavity and pack gently to bring the surface of the powder up to the reference plane Observations Randomization The powder must be sufficiently compact to provide cohesion without using a binding agent or needing to smooth the surface. Too much pressure can cause orientation. It is thus necessary to exploit the degree of “randomization–orientation” by preparing the powders as regularly as possible. Indeed, as the clay layers are planes, they tend to be oriented, which is likely to give irregular results. This effect can be limited by using a binding agent that is inactive both with respect to the X-ray and the sample (acetone, lake gum + alcohol, collodion + acetate of butyl or amyl, gum tragacanthe, etc.). These treatments make the powders more stable. However, the use of a vertical goniometer prevents separation of the powder when it is placed in the beam which makes this kind of preparation unnecessary in the majority of the cases.

a b
Fig 4.4. Supports hollowed out for diffraction: plastic support (a), Siemens revolving support (b), hollowed glass support (c)

c

When only a small quantity of sample is available, double-sided adhesive Scotch tape can be used; this is powdered with the sample (the surplus can be removed by light tapping) or better still, by placing the sample on a silicon support. Freeze-drying makes it possible to obtain less oriented samples (although there is a risk of a few packages of layers displaying residual orientation). Rotating the sample enables maximum possible reflection,

X-ray Diffractometry

93

which decreases the risk of error by improving randomization and by limiting fluctuations in intensity due to an insufficient number of particles. The rotating support requires larger quantities of powder. Granulometry – Focusing The use of a sample plane is required for satisfactory focusing of the beam. The roughness of the surface has a marked effect on the relative intensities of the lines. If the surface is rough, as is true in the case of a coarse powder, the absorption coefficient will be high and the intensities at small angles will consequently be exceptionally low. The powder must be fine (a particle size of about 10 µm) to avoid such fluctuations in absorption, but not too finely ground to avoid an artificial increase in amorphous minerals: – at 10 µm, fluctuations will not exceed 2% – at 50 µm, fluctuations can reach 20%. On the other hand, with a particle size lower than 0.02 µm, diffractions are diffuse and the intensity decreases. The width of the diffraction curve increases with a decrease in the thickness of the crystal. The structure of “amorphous” substances is characterized by the absence of periodicity or by short-range organization. In the latter case, they show only one statistical preference for an interatomic distance and XRD cannot give satisfactory results but only broad and badly defined peaks (Fig. 4.5).

Fig. 4.5. Diffraction diagrams typical of (a) a well crystallized substance and (b) an amorphous substance with shortrange organization

Particular uses of Powder Diagrams A powder diagram can be used to identify non-transformed primary minerals (quartz, calcite, etc.) or oxides and hydroxides of iron and

94

Mineralogical Analysis

aluminium, etc, as the degree of crystallinity in the case of kaolinites and the “fire-clays” will be distinguished more clearly: since the mode of stacking of the layers is different, certain peaks of kaolinite do not exist in “fire-clays”, which may consequently not be seen on a oriented diagram. The di- or tri-octahedral nature of the layers can be highlighted by using ray 060 [filling of octahedral cavities by bivalent (1.54 Å) or trivalent (1.49 Å) cations]. Polymorphous illites can provide data that are characteristic of the mode of formation. 4.2.3 Preparation for Oriented Diagrams

Oriented Diagrams on Glass Slides Objective The objective is to identify certain clayey minerals capable of being oriented and to observe basal variations by means of heat or chemical treatments. The number of treatments needed depends on the mixture of minerals in the sample. Principle In diagrams of oriented aggregates, special weight is given to the crystalline planes parallel to the surface of the layers. The preferential orientation of silicates causes an increase in the maximum of basal differentiation d(00l), which makes it possible to detect small quantities of crystalline species present in the mixture. With this method better diffraction is obtained of the species and fluctuations that are theoretically lower than with randomized powders since the particles do not exceed 2 µm. On the other hand, preferential orientation decreases the number of planes (hkl) in a position of diffraction. Procedure (a) Starting from a well-homogenized clay paste obtained by centrifugation, remove a small aliquot (approximately 400 mg) with the spatula and place it in a tube with 5 mL water; agitate to suspend particles. (b) Starting from a powder (for example recovery of the clay used in Chap. 9), weigh approximately 200 mg of clay, place it in a plastic tube

X-ray Diffractometry

95

with 5 mL water and a 8 mm glass ball and agitate for 2–3 min to disperse the clay. (c) Starting from a or b, pipette approximately 1 mL of suspension and spread it evenly on a 24 × 24 mm glass slide. The deposit must be spread evenly over the whole surface of the slide with no areas of extra thickness. If the deposit is too thin, there may be an effect of the support, if it is too thick, diffraction will occur only on the finest clay and the clay film may reticulate itself. Allow to dry at room temperature, then dry in the desiccator. Prepare three slides: – the first for examination of the rough sample without treatment – the second for examination of the sample after glycerol or glycol treatment – the third for heating the sample in the oven at 490°C (depending on requirements, this slide can be used for several thermal treatments at increasing temperatures). Observations After the suspension has been spread on the glass slide, microfractionation will occur as a function of the size of the clay particles since coarse clay sediments faster than fine clay. However, if drying is rapid, this segregation will not disturb qualitative interpretation. In contrast to powder techniques (cf. Sect. 4.2.2) that have to be sufficiently thick to limit the effects of orientation, oriented slides must be sufficiently thin so that the maximum number of basic units are suitably oriented. Plotting a black line on the support before use is a good way to judge the quality of the preparation: it should remain very slightly visible through the almost transparent clay film. In comparative semi-quantitative analysis, the suspended deposit of each sample should contain about the same quantity of clay. This is easy starting from dry clay which can be weighed before suspension. The removal of an aliquot of the sample during agitation results in slides with almost the same quantity of mineral material spread over the same area. From a suspension after agitation and homogenization, quickly remove an aliquot of the same volume and dry and weigh one of the samples to determine the concentration. For suspensions of wet materials, recover clay on a film of a given surface area and weigh an identical aliquot after drying. Amorphous minerals can mask part of the diagram. It is often necessary to eliminate them before recording the diffraction diagram.

96

Mineralogical Analysis

Diagram on Oriented Paste Principle This method is particularly suitable for minerals of the 2:1 type and halloysite–4H2O, the wet sample should be dried at room temperature and maintained at a relative humidity of approximately 80%. Procedure After insulation of clay to the required dimension by centrifugation (cf. Chap. 3), recover the wet centrifugation pellet, homogenize with a small stainless steel spatula, deposit it on a hollowed support (Fig. 4a), spread it out in the cavity, then smooth it with a glass slide to a perfect plane suitably located on the reference plane. Allow to dry slowly at ambient temperature taking care to avoid excessive desiccation. Observations On coarse 2 µm clays, the clay paste needs to be homogenized after centrifugation, as the part near the surface is able to concentrate fibrous clays which do not respect Stokes law. Smectites deposit preferentially at the surface whereas chlorites and illites sediment more quickly at depth and accumulate more extensively at the base of the centrifugation pellet. The same is true for iron oxides (density 4.5–5) which are heavier than aluminium oxides (density approximately 3). During drying, clays with a high coefficient of retraction can fissure and detach. A binding agent can be added, though there is a risk of introducing a variable that is not easy to control (disturbed orientation, flocculation, binding agent not amorphous for X-ray, etc). Specific Uses In certain cases, it is possible to work on the wet sample. With 2:1 clays, saturating the wet paste with Na+ or Li+ makes it possible to observe the phenomenon of unlimited swelling d(00l) > 30Å. During air drying, clay again reaches values of d (00l) = 18, 15, 14 Å. Aggregates Oriented on Porous Ceramic Plate Objective To allow retention of minerals that do not adhere on glass slides and/or, to allow successive treatments on the same sample: original sample, cation treatment, polyalcohol treatment, followed by successive heat treatments. This preparation enables satisfactory orientation of clays, and is particularly useful if the spectra are to be exploited for quantitative analysis.

X-ray Diffractometry

97

Principle The clay is fixed on a porous plate by suction using a vacuum pump or by centrifugation using a Poretics ceramic porous disc and a Hettich support. Procedure Transfer the clay suspension on a 24 × 24 mm porous plate placed on a Büchner with a diameter of 40 mm whose bottom of sintered glass has been modified to create a 22 × 22 mm window, which is the only permeable part (Fig. 4.6). Tip the homogenized liquid onto the plate and apply a vacuum using a standard pallet pump. After formation of a thin continuous layer of clay, dry the sample and analyse by X-ray diffraction. Observations This medium and procedure gives well-oriented deposits with a reduced fractal dimension of the surface which is preserved after drying and is thus of excellent quality for XRD measurement. The thickness of the deposit must be homogeneous all over the plate and sufficiently thick to avoid the effects of the support, but thin enough to preserve a certain degree of elasticity and to avoid cracking caused by the rheological properties of the medium.

Fig. 4.6. Preparation on porous ceramic plate

Measurements on porous plate make it possible to avoid excess solvation liquid and also to carry out the treatments required: chemical saturation by Mg2+ or K+ treatment, heating, etc. 0.20 µm millipore filters on glass plates can also be used for measurements (collection of suspended materials in water, airborne dust, etc). The cleaning of the porous plate is delicate and should be done after washing by slight abrasion of the surface that has been in contact with the

98

Mineralogical Analysis

clay. It is important to preserve the perfect flatness of the plate and to make sure contamination has not occurred. Oriented Aggregates Deposited by Ultracentrifugation On Semi-Flexible Film Objective The aim of this method is to obtain samples of the same thickness, the same mineralogical composition and the same apparent particle size. Principle Particles are deposited by means of a Sharples super-centrifugal machine equipped with a cylinder with a semi-rigid internal plastic film (cf. Chap. 3). The speed is regulated to collect particles of a known average diameter (coarse to fine clay). If necessary, the film can be analyzed at eight successive levels of 24 mm, which makes it possible to monitor modifications in the nature of clays as a function of particle size and density. To ensure the film is not too thick, the proportion of coarse, medium or fine clay must be known, and only the zones that correspond to an optimal density used (test of the black line on the film, cf. section “Oriented Diagrams on Glass Slides).

Fig. 4.7. Diagram showing how to cut up a plastic film after continuous flow ultracentrifuge and how to set it up for XRD.

Procedure Remove the plastic film from the cylindrical bowl and dry flat for approximately 1 h. When it is still slightly wet and not yet rigid, cut it

X-ray Diffractometry

99

into sections following the 24 × 24 mm pattern drawn on the back of the film (Fig. 4.7). Each level should be labelled. Assemble the 24 × 24 mm semi-rigid film on a glass plate and place it on a support against a stop corresponding to the reference plane. Observations XRD can be performed directly on the oriented sample or after treatments (cf. Sect. 4.2.4) with polyalcohol; heating treatment at 110°C is also possible depending on the nature of the plastic used. The surface of the plastic support must be unpolished naturally without additives as these could cause background noise that is incompatible with the smoothness of measurements. 4.2.4 Pretreatment of Clays

Effect of the Pretreatments Different types of pretreatments can be used to facilitate the identification of clayey minerals by causing selective changes in the inter-reticular distances of the clayey layers. These treatments and their effects are summarized in Table 4.4. Table 4.5 shows the maximum inter-reticular distances observable in the clayey minerals of soils. Saturation by Cations Principle Saturation by a cation makes it possible to fill the existing cation vacancies and, by displacement of the exchangeable cations, to obtain homoionic samples that present uniform expansion of the layers of the expansible phyllosilicates (the quantities of interlayer water depend on the exchangeable cations). The divalent cations, e.g. Mg2+, Ca2+, Sr2+, give hydrates with two layers that are relatively stable in a broad range of relative humidity and are not very sensitive to the influence of hydronium ions during washing with water. Mg2+ saturation gives a stable complex with two layers of inter-layer water which brings the expansible phyllosilicates to d 001 ≥ 14 . Expansion of the layers allows differentiation of the non-expansible varieties of clays whose interlayer space is approximately 10 Å (Table 4.5). K+ saturation restricts interlayer adsorption of water and allows differentiation of 2:1:1 clays of the chlorite type, and of 2:1 clays of the vermiculite type. Non-expansible chlorites are not modified by this

100

Mineralogical Analysis

treatment, whereas vermiculites break down at 10 Å. The same sample can be also used for smectites after heating to 550°C. Li+ saturation followed by dehydration to approximately 300°C, followed by solvation with glycerol, makes it possible to differentiate montmorillonite from beidellite using the Greene–Kelly test based on the Hofmann–Klemen effect. When a smectite saturated with Li is dehydrated at 300°C, the Li interlayer migrates towards the octahedral layers which have a deficit of positive charges resulting from substitutions. The structure becomes non-expansible and there is no further inflation of the sample with glycerol treatment (9.5Å). A beidellite (or a saponite) whose charges comes from tetraedric substitutions is not affected by such treatments and inflates with glycerol (17.7 Å). Procedure for Mg2+ saturation – Take an aliquot of 125–250 mg of clay – put in suspension and acidify with diluted hydrochloric acid to bring it to pH 3.5–4.0 to avoid precipitation of magnesium hydroxide (particularly if initial dispersion was carried out in an alkaline medium) – add 5 mol L −1 magnesium acetate solution to obtain a suspension of approximately 0.5 mol (Mg) L−1 – leave in contact for 30 min – centrifuge for 5 min at approximately 3,000g and discard the supernatant – wash the centrifugation pellet twice with 0.5 mol (Mg(OAc)2) L−1 solution to eliminate H+ from the acid suspension then twice with 0.5 mol (MgCl2 ) L−1 solution(approximately 10 mL) – centrifuge and wash with 50 % methanol (approximately 10 mL) then with 98% methanol (approximately 10 mL); then with 85 % acetone. − The silver nitrate test for Cl must be negative – dry at room temperature for the powder spectrum, or prepare the plates for oriented spectrum immediately by adding a little water. Procedure for K+ saturation – Take an aliquot of 100–250 mg of clays – put in suspension and add 1 mol (KCl) L −1 solution – centrifuge the flocculated clay after 30 min of contact – discard the supernatant – wash the centrifugation pellet with KCl 1 M

Table 4.4. Influence of the treatments on the inter-reticular distance between layers of principal clays: air dried samples, saturated with Mg2+ (M = metahalloysite or halloysite,2H2O (7 Å), H: hydrated halloysite, 4H 2O (10 Å). Intergrades of chlorite–vermiculite and chlorite–smectite are also separated by these treatments

X-ray Diffractometry

group 9.4 – talc micas pyrophryllite (illite) sepiolite palygorskite 0.2–0.6 expansion 17.7 Å 0.6–0.9 nil 14 - 15 Å smectites vermiculites 10 cation 10–12 fibrous 14–15 water cations

1:1 (Te – Oc)

2:1 (Te – Oc – Te)

2:1:1 (Te – Oc– Te – Oc octahedral chlorites

d (Å) Stability (inter-layer) Family

7 –

7–10 H2O

kaolinite

serpentine M halloysite H 1–2 nil 10.1 Å nil 10.1 Å nil 10 Å

Charges x Glycerol solvation

0 nil 7.1 Å

nil 7.3 Å

variable nil except swelling of chlorites

K saturation

+

nil 7.1 Å

nil 7.3 Å

Heating 550°C

collapse

nil 7.3 Å

0 nil nil M 7.2 Å H 10.8 Å nil nil M 7.2 Å H 10 Å M collapse nil H collapse nil S 12.1 Å P 10,48 Å nil S 12.1 Å P 10.48 Å 10,4–8,2 Å 9.2 Å 4.7 Å contraction 12.4 Å 12.8 Å contraction 10 Å

contraction nil 10 Å (avec 14 14 Å Å) contraction nil 14 Å

Formamide intercalation DMSO

nil 7.1 Å <4h expansion 11.18 Å

nil 7.3 Å nil 7.3 Å

expansion M 10.48 Å M expansion montmorillonite 9.5 Å beidellite 17.7 Å

Green–Kelly test, 250°C glycerol

101

102

Mineralogical Analysis
2+

Table 4.5. Maximum inter-reticular distance of soil minerals saturated with Mg and K+ or solvated by glycerol treatment. mineral (a) Samples saturated by Mg2+ chlorite vermiculite montmorillonite mica (Illite) talc halloysite metahalloysite kaolinite lepidoscrocite boehmite gibbsite silicates gypsum göethite cristobalite lllmenite quartz feldspar calcite hematite magnetite trioctaedric silicates dioctaedric silicates 13.6 – 14.7 14.0 – 15.0 14.0 – 15.0 9.9 – 10.1 9.2 – 9.4 10.1 7.2 – 7.5 7.1 – 7.2 6.27 6.11 4.85 4.4 – 4.6 4.27 4.18 4.04 3.73 3.34 3.1 – 3.25 3.03 2.69 2.53 1.54 100 104 311 060 001 002 001 001 002 001 001 001 020 020 002 110 121 110 101 102 101 d (Å) hkl plane

1.49 2+ and solvated by glycerol (b) Samples saturated by Mg montmorillonite vermiculite chlorite halloysite montmorillonite (2nd order) chlorite – Vermiculite (2nd order) 17.7 14.4 13.6 – 14.7 10.1 – 10.7 9.5 7.15 001 002 001 001 001 002

minerals giving inter-layer spaces similar to ‘a’

X-ray Diffractometry
(c) Samples saturated by K+ and air dried chlorite montmorillonite vermiculite metahalloysite chlorite (2nd order) 13.6 – 14.7 11.0 – 13.0 10.0 – 11.0 7.2 – 7.5 7.15 001 001 002 001 002

103

Minerals giving inter-layer spaces similar to ‘a’ d) Samples saturated by K+ and heated at 550°C for 2-3 h chlorite montmorillonite vermiculite mica (Illite) chlorite (2nd order) 13.6 – 14.7 9.9 – 10.1 9.9 – 10.1 9.9 – 10.1 7.15 001 001 002 001 002

– then wash with 50% and 95% methanol and finally with 95% acetone until there are no Cl− ions in the washing solution (no precipitate with addition of AgNO3) – allow to dry at room temperature or immediately prepare slides for oriented spectrum by adding distilled water. Procedure for Li+ saturation – Take an aliquot of 110 – 250 mg of clay – put in suspension and add 3 mol (LiCl) L−1 solution – leave in contact for 30 min – centrifuge and discard the supernatant – rapidly wash the centrifugation pellet with 1 mol (LiCl) L−1 solution then with a little water – make an oriented slide and dry at 250°C overnight – carry out the glycerol treatment (1 night) and perform the diffraction spectrum. Remarks – This test is not completely selective – mounting on an ordinary glass slide can cause errors during heating as the Na+ in glass can exchange with the Li+ in the clay, which causes incomplete neutralization of the octahedral charges – the glycerol treatment must be carried out hot for a period of several hours to allow complete expansion of the layers – an irrational basal sequence indicates inter-stratification.

104

Mineralogical Analysis

Removal of Iron Principle Iron often has to be removed first to mitigate its action on the process of measurement by XRD using a Cu tube (X-ray fluorescence increases background noise) and second to avoid dilution through a reduction in the intensity of diffraction. Different methods can be used to eliminate the different forms of iron. These methods vary in vigour and should not transform the phyllosilicates present, but make it possible to complex and reduce amorphous and crystallized iron in a slightly acid medium with the minimum possible degree of aggressiveness. These methods are similar to those used in Chaps 2 and 6, but the solubilization of iron compounds is generally not controlled (because it is the final residue of the sample will be analyzed by XRD). It should be noted that the amorphous silicon–iron complexes present in certain sediments will be dissociated in an acid medium giving soluble iron and precipitated amorphous silica. Hematite and goethite oxides are only slightly affected by an acid oxalate treatment at pH 3. A reducing treatment with complexation of the products of dissolution (oxalate acid + dithionite) is the best way to eliminate most iron while sparing the clay. Oxalate dissolves amorphous iron and dithionite dissolves oxide forms of iron. A slightly acid medium allows extraction of “free iron” but at the same time extracts part of the iron of the lattice of certain clays, for example vermiculites. With dithionite, the presence of sulphur precipitated after reduction does not pose a problem for XRD, except for a dilution effect on the sample. Sulphur should be eliminated from the extraction pellet after centrifugation and drying. Procedure The main methods are based on dissolution in an acid or base sequestering medium and/or reducing medium (cf. Chap. 6): – The DCB method (dithionite, citrate, bicarbonate) extracts iron from the majority of its amorphous and crystalline compounds by reduction and sequestration without significant modification of aluminosilicates or lithogenic hematite – the method based on acid oxalate at pH 3.0 (in darkness, or with UV photolysis) extracts noncrystalline forms – the sodium pyrophosphate method at pH 9–10, EDTA at pH 9–10, acetyl acetone extracts organometallic forms

X-ray Diffractometry

105

– the tetraborate method at pH 8–9 extracts Fe from monomeric complexes. The elimination of iron is accompanied by the dissolution of aluminous products, silica, etc. which can be controlled by chemical titration of the extracts. Solvation Treatments Principle Solvation by polar molecules, such as mono or polyhydric alcohols, ethers, amines, results in the formation of interlayer organic complexes. The resulting structure is more stable than the structure of dehydrated 2:1 clay. Swelling is all the easier since the charge of the layers is weaker, or is limited to the octahedral layer. The nature of the interlayer cations modifies the limits of stability of the organic complex. The basal distance of the smectites reaches 17.7 Å with a double interlayer layer of glycerol, and 17.1 Å with ethylene glycol. The rate of hydration can vary considerably. Montmorillonites inflate more easily than the majority of clays. Impregnation can be accomplished in the liquid or vapour phase by heating to 60°C. In certain cases, the treatment has to be continued for 24 hours to take the slowness of interlayer expansion into account. Condensation of the vapour does not cause any mechanical disturbance and gives more intense lines of diffraction. Procedure for Glycerol Treatment This procedure is based on that of Modre and Dixon (1970): – Prepare a 1:10 mixture of glycerol and water – on a previously air-dried oriented plate, apply a film of glycerol in a very fine spray, taking care not to create an excess of reagent – allow to dry for at least 1 or 2 h and then perform XRD. Caution. The complex loses its effectiveness over time, so it is advisable not to wait more than 20 h. Using ceramic plate for this treatment makes it possible to eliminate excess glycerol.

106

Mineralogical Analysis

Procedure for Ethylene Glycol Treatment

Fig. 4.8. Treatment of the samples with ethylene glycol

This procedure is based on that of Eltantawy and Arnold (1974) and Chassin (1974): – place the sample (generally for spectrum-oriented aggregates) in a desiccator containing the ethylene glycol (Fig. 4.8) – create a partial vacuum with the vacuum pump and leave the samples in contact with the vapour phase for at least one night – remove the sample and perform XRD as soon as possible. Caution. The complex loses its effectiveness over time, so it is advisable not to wait more than 10 h. Safety. Ethylene glycol or 1,2 ethanediol: HOCH2–CH2OH, has a boiling point of 196°C; it is hygroscopic, toxic by ingestion, and can affect the kidneys, lungs and the heart. Intercalation Complexes Principle Intercalation complexes are very useful particularly to separate 1:1 clays and to distinguish well-crystallized forms, disordered forms and halloysites. Since the forms of these species are similar, it is impossible to separate them with precision (kaolinite and halloysite can be found in flat, tubular or glomerular forms). One common procedure is first to form the intercalation complex, then after obtaining the spectrum, to move the complex with water and perform solvation with ethylene glycol or glycerol. The stability of the intercalation complexes is variable and XRD spectra should be performed without delay.

X-ray Diffractometry

107

Treatment with hydrazine hydrate (Wada and Yamada, 1968; Range et al. 1969; Calvert, 1984) allows the inter-layers of the well-ordered kaolinite to be changed from 7.15 to 10.48Å without modifying the chlorite. The presence of interstratified minerals can be awkward. In the treatment using dimethylsulfoxide (Gonzalez Garcia and Sanchez Camazano 1968 Olejnik et al., 1968 Anton and Rouxhet, 1977; Calvert 1984): (i) kaolinite forms an intercalation complex that increases the interlayer distance from 7.15 to 11.18 Å, which remains stable after heating to 300°C, (ii) halloysite and dickite display identical behaviour, except that heating to 300°C does not result in further expansion, (iii) vermiculites and smectites increase from 18 to 19 Å and (iv) chlorite undergoes no change. Metahalloysite forms an intercalation complex with formamide. The line rapidly reaches 10.4 Å and whereas for kaolinite there is no reaction even after 4 h of contact (Churchman et al., 1984). Procedure for Hydrazine Hydrate Treatment – Place the clay sample on a glass slide or on porous ceramic in a saturator containing the hydrazine hydrate – create a partial vacuum with the vacuum pump and leave the sample in contact with the vapour phase at 65°C for at least one night – remove the sample and perform XRD without delay. Remarks It should be noted that the complex loses effectiveness over time; so it is advisable not to wait more than 1 h before XRD. Sequential treatment with hydrazine + water + glycerol makes it possible to increase the basal interlayer distance of halloysite to 11.1 Å, whereas that of kaolinite remains at 7.15 Å. Safety – Hydrazine hydrate: H2NR–NH2,H2O, boiling point 119°C – miscible with water and ethanol – strong base, very corrosive, attacks glass and the skin – highly toxic, causes irritation of the eyes, may be carcinogenic. Procedure for Dimethylsulfoxide Treatment – Put an aliquot of clay weighing from 100 to 250 mg in suspension in 5 mL of dimethylsulfoxide – leave in contact for 8 h in a water bath at approximately 40°C – agitate from time to time – centrifuge and prepare slide, allow to dry – rapidly perform XRD after drying.

108
Table 4.6.

Mineralogical Analysis
Effect of the thermal treatment on the diffraction of clays (TDA = characteristic temperature of change in thermal differential analysis, cf. Chap. 7).
temp. (°C)

mineral

4 H TDA at

heat effect absence of diffraction

Well crystallized Kaolinite Disordered kaolinite 1:1 clays Dickite Nacrite Halloysite, 4H2O Metahalloysite, 2H2O Shortrange order aluminosilicates Crystallized micas (muscovite)

500

575–625

500 500 500 110 500 110 500

550–562 665–700 625–680 125–160 560–605 125–150 560–590 140–180

absence of diffraction absence of diffraction absence of diffraction elimination of water absence of diffraction elimination of water absence of diffraction elimination of water, no XRD spectrum difficult identification by XRD →18Å

Allophane Imogolite

110 350 500 490 350

500 700 125–250 350–550 700 530–650 500–600 700

disappearance of 18Å peak, destruction of the mesh mica spectrum up to 1,000°C loss of water mica spectrum mica spectrum (001) loss of water–mica structure mica structure mica spectrum until 700– 1,000°C Progressive loss of H2O the initial basal space (001) is a function of moisture: 14– 13, 8–11, 6–9Å changes Disappearance 15 → 9Å

Illite–micaceous clays 500 500 Micas and 2:1 Glauconite Celadonite Biotite 500 500 500

clays

Vermiculite

300

300

Montmorillonite

300

300

X-ray Diffractometry
Chlorite 500 680–800

109
no change if structure is well ordered intensification of line 14 Å – line 3.54 Å not affected (octahedral layer 820°C intercalated layer 640°C)
Attenuation of line 14Å which

Chlorite–Mg

500

650

2:1:1

becomes broad and diffuse. octahedral layer 530°C, Fe2+

Clays

Chlorite–Fe

500

500

intercalated layer 430°C,
Fe2+

intercalated layer 250°C
Fe3+ (octahedral layer 750°C Chlorite–Al Inter-stratified 500 500 < 600 intercalated 900°C layer) effects vary with the mineral species present <200 dehydration space 12 Å becomes weak and diffuse → 9.8 Å, space 7.6 Å more intense recrystallization dehydration without change in structure, 10.5 Å peak becomes broad and diffuse, destruction of structure

Fibrous

Sepiolite

350

>200 350 800

Clays

Palygorskite

300 500

<400 440 800

Safety – Dimethylsulfoxide: C2H6OS, boiling point: 189°C – hygroscopic, irritation of the skin (urticant), keep away from the eyes. Procedure for Formamide Treatment – Place the clay sample on a glass slide or preferably on porous ceramic, allow to dry – vaporize formamide and note the time of application; when the formamide excess has been eliminated (approximately 20 min), perform XRD (repeat with the same sample at the end of the day and compare the two spectra).

110

Mineralogical Analysis

Safety Formamide (H–CO–NH2) is an ionizing solvent which can release a slight odour of ammonia. It dissolves lignin. No known risk to health. Thermal Pretreatments Principle The hydrated minerals undergo modifications due to the effect of the rise in temperature. These modifications occur at certain characteristic temperature stages. The length of time at a given temperature is also significant. The transformations take place with varying rapidity depending on the nature and the degree of crystallinity of the thermally sensitive minerals. In general, 2–4 h are required
Table 4.7 Effect of thermal treatments on oxides and hydroxides of aluminium and iron (TDA: characteristic temperature of change in thermal differential analysis, cf. Chap. 7) mineral temperature (°C) 4H TDA iron series Goethite 300 900 Goethite alumineuse Lepidocrocite 300 350 230–280 spectrum of disordered hematite 900 spectrum of well-crystallized hematite heating effect

240–350 spectrum of disordered hematite 900 spectrum of well-crystallized hematite → hematite spectrum → hematite spectrum → hematite spectrum At 300°, the akaganeite spectrum weakens 230–280 γ Fe2O3 broad peak 400–500

Maghemite Magnetite

350 500

350–450 600–800

Akaganeite δFeOOH Feroxyhyte δFeOOH Ferrihydrite (without Si)

300

200–400 gradually 420–500 → hematite spectrum → hematite spectrum with intermediary goethite unstable in air at ambient temperature→ goethite spectrum after air drying → hematite spectrum

300

140–260

350

350–400

X-ray Diffractometry
Ferrihydrite (with Si) Aluminium series Gibbsite Bayerite Nordstrandite Boehmite Diaspore 200 200 200 500 500 150–200 150–200 150–200 450–500 470–500 → boehmite and γAl2O3 spectrum/a? → boehmite and γAl2O3 spectrum/a? → bayerite spectrum → boehmite → γAl2O3 spectrum 500 550–600 → hematite spectrum

111

→ spectrum of disordered corundum towards spectrum of crystallized corundum

Procedure Place the samples assembled on glass slides in a cold furnace; bring the furnace to the desired temperature (110, 350, 490, 530°C, etc.) and maintain the temperature for at least 4 h. Heating must be progressive to avoid breaking the glass slides and possible reticulation of the clay film. The furnace should then be allowed to cool gradually, opening the door to accelerate the process if necessary. If they have not undergone irreversible transformations, the slides can be stored in the desiccator until XRD. Table 4.6 shows the influence of heat treatments on the diffraction of clays. Table 4.7 shows the influence of heat on the diffraction of aluminium oxides and iron hydroxides. Observations – Loss of interlayer water results in contraction of the mesh and displacement of the basic diffraction lines – heating to high temperatures can lead to collapse of the lattice and dispersal of the characteristic X-ray spectrum. Example Dissociation of kaolinite by heating (Fig. 4.9) to around 490°C can be visualised by the four following reactions: (1) at 500°C: Al2O3,2SiO2,2H2O → Al2O3,2SiO2 (metakaolinite) + 2 H2O (2) at 925°C: Al2O3,2SiO2,2H2O → Al2O3,SiO2 (sillimanite) + SiO 2 (ß quartz) + 2 H2O (3) at 1 100°C: Al2O3,2SiO2,2H2O → αAl2O3 + 2SiO2 (ß quartz) + 2H2O (4) at 1 400°C: Al2O3,2SiO2,2H2O → 1/3 (3Al2O3,2SiO2) + 4/3 SiO2 (ß quartz) + 2H2O

112

Mineralogical Analysis

All these reactions are possible starting from 800 K (approximately 527°C). There is no obstacle to the transition of kaolinite → metakaolinite → sillimanite → mullites → oxides, the most stable system, which depends on the temperature. Thermodynamically reaction (1) occurs first (Fig. 4.9).

Fig. 4.9. Transformation of kaolinite by heating

Preparation of Iron Oxides for XRD Principle The use of XRD to study iron oxides in the soil requires a sufficient concentration of the different iron phases. Consequently the products have to be concentrated and purified without modifying either their crystallinity or the level of substitution by aluminium, and without causing chemical conversions of the phases. The following methods can be used to this end: – separation methods (concretions, separation by density gradient, magnetic separation, etc – in situ XRD determination on an uncovered thin slide, or extracted micro-samples with high iron content. Chemical methods only enable the concentration of iron, which can exist in the form of coating, in amorphous forms (ferrihydrite, etc.), or involve varying degrees of crystallization (goethite, hematite,

X-ray Diffractometry

113

lepidocrocite, etc.). Clay minerals are dissolved by 5 mol (NaOH) L−1 solution under boiling. Kaolinites, halloysite, gibbsite, amorphous aluminosilicates, etc. are destroyed and solubilized, but 2:1 clays are more resistant and are thus only partially attacked. In addition to iron, smectites, illite, quartz, anatase, and rutile are also found in the residue. If the amount of silica present is not sufficient, during heating with 5 mol (NaOH) L−1 solution, the ferrihydrites are likely to be transformed into hematite or goethite. If the sample is attacked using 5 mol (NaOH) L−1 + 0.2 mol (silica) L−1 solution, the rate of aluminium substitution is not significantly modified. Procedure – Weigh in PTFE beakers 2–5 g of soil, or granulometric clay <2 µm depending on the total iron content – attack with 20 mL 5 mol (NaOH) L−1 + 0.2 mol (silica) L−1 solution, cover the beaker with a PTFE plate and boil for one hour – allow to cool and decant – dilute the medium with deionised water, then wash on filter with water until the pH is neutral – perform XRD Note: a powder diagram is better than an oriented diagram for these products. Remarks – If the soil or clay contains many 2:1 phyllosilicates, the concentration may be too low; in this case it is useful to perform two differential spectra (DXRD), one with the original product, the other with the chemically concentrated product, or with the spectrum obtained after dissolution of iron by the citrate–bicarbonate–dithionite mixture (CBD, cf. Chap. 6) – a cobalt X-ray tube should be used to avoid fluorescence of iron which occurs with a copper tube – the PTFE beakers can be used up to a maximum temperature of 250°C; – in certain cases, destruction of the matrix silicates can be achieved through an HF attack in a PTFE beaker – heating to 600°C gives anhydrous oxides → hematite. 4.2.5 Qualitative Diffractometry The sample of whole soil crushed to 0.1 mm and “clay < 2 µm” fractions (or 0.5 or 0.2 µm sub-fractions,) prepared for randomized or oriented powder spectrum are analyzed by XRD.

114

Mineralogical Analysis

Each group of clay has its own layered structure which gives basal reflections whose position, intensity and shape enable either immediate identification or after specific treatments. This section describes the interpretation of standard situations that make it possible to evaluate clayey minerals at the scale of the group and the sub-group, and very occasionally at the scale of the species. Fine Adjustment of Experimental Conditions The experimental conditions must be precisely determined in advance, i.e. the type of tube (Cu, Co, etc.), the intensity applied to the filament, the opening of the slits, the speed of rotation of the goniometer, tension meter, amplification, adjustment of the constants of time–inertia, sensitivity, signal/background ratio, angular zone, scanning speed, etc. A cobalt tube X-ray should be used for samples that are rich in iron because it does not generate fluorescence with this ion. Generally, with suitable geometry, an international standard Cu tube give a satisfactory performance, and JCPDS tables and software (see note 1 in Sect. 4.2.2) cited here are based on this standard. The rotation speed of the goniometer and the selection of the angular zone of scanning determine the time needed for analysis. The choice of the scanning rate of the sample influences the relative precision of the diffractograms. Too high a speed can lead to insufficient discrimination. For the determination standards, a displacement speed of 1 to 2 °min−1 (2θ angle) may be sufficiently rapid but does not allow adequate separation of fine doublets such as those of kaolinite (002) and chlorite (004) for which a slower rate is necessary (<0.5° 2θ min−1). For rapid characterization of clay (group and sub-group) the angular zone of scanning can be limited to a 2θ zone of approximately 2 – 40 degrees, if determination of the d(060) diffraction is not required as this takes much longer. The strong reflection at 2θ = 60° makes it possible to estimate parameter “b” and thus to differentiate dioctahedral minerals (1.48–1.50 Å) from trioctahedral minerals (1.53 – 1.55 Å) and sometimes to provide more evidence for approximate values by bringing these values closer to the intensities of d(001) diffractions, for example in the identification of certain micas. Powder samples are performed by rotating the samplesupport in the reference plane, which requires high precision equipment. The oriented sample method makes it possible to reinforce the reflections (001) by directing the particles according to the development plane of clay minerals. All the slides (or any other form of support) for each sample should be made simultaneously and as homogeneously as possible

X-ray Diffractometry

115

(same suspension, same quantity of clay per slide, etc.). Three subsamples are usually essential: – a reference oriented sub-sample without treatment, dried and observed with a known quantity of relative moisture – an oriented sub-sample solvated for hydrated 2:1 clays – an oriented sub-sample for heat treatment (500°C for 4 h). The samples should be analysed one after another, and the equipment should be regulated in exactly the same way for each sample to enable better comparison of the spectra. In more complex cases, it may be necessary to prepare homoionically saturated samples (K+, Mg2+ ) with suitable treatment and intercalation sequences. These should be performed under the same conditions taking into account the relative humidity of the air, relative stability of the solvations and complexes of intercalation, reaction times (in the case of formamide treatment whose action is rapid on halloysite, whereas on kaolinite this treatment can take 4 or 5 h and should thus be performed at the head of the series). If the sample is very small, it may be possible to perform treatments and measurements sequentially on the same sample. In this case porous ceramics should be used (or 0.20 µm Millipore filters if heat treatments are not envisaged). However, as measurements must then be made over a period of several days it is more difficult to maintain constant conditions. Examination of the Diagrams Background Noise The background noise is due to non-specific signals that are inherent to the material, to the process, to minerals with short-range organization, and possibly to fluorescence phenomena resulting from the emission of secondary X-rays. The latter phenomenon can be eliminated by the back monochromator and amplitude discrimination. The background noise can be also smoothed electronically with suitable software. In the zone of the small angles (2θ between 0 and 10° approximately), depending on the adjustment of the slits, the counter may receive part of the direct beam or of reflections from the edges of the support. The use of variable slits generally prevents these phenomena. The presence of amorphous substances (aluminosilicates, different oxides, etc.) is indicated by broad and diffuse bands. The interpretation software makes it possible to eliminate these zones selectively and to stabilize the base line.

116

Mineralogical Analysis

Geometry of the Peaks The geometry of the peaks enables determination of the degree of crystallinity, particularly for well-crystallized minerals where crystal size is suited to the X-ray wavelength and presents regular inter-reticular variations giving clear diffractions and many harmonics (Fig. 4.5a). Badly crystallized minerals that are not well ordered and whose size is not suitable to X-ray often give broad biconvex peaks whose surfaces are not usable for quantitative analysis (e.g. smectite, Goethite, see Fig. 4.5b). Too many particles such as diatoms, quartz or feldspars, can limit the orientation of clays and render interpretation of the spectra difficult. Superposition of peaks resulting from the cumulative presence of several minerals results in a widening of the signals and an abnormal rise in relative intensities, for example Chlorite (001)–Vermiculite or Kaolinite (001)–Chlorite (002). These samples require different pretreatments to reduce uncertainty and to reveal the masked peaks (cf. Sect. 4.2.4). Peaks may display low intensity and widening even in the case of minerals that are usually well distinguished such as quartz, calcite or ordered kaolinite (001 or 002) if the sample is not thick enough. Goethites and smectites which have more diffused peaks may be masked. At certain angles, lines that come from the support are also likely to appear. Particles of approximately 0.5 µm are used for comparative studies on different clayey fractions and for quantitative analysis, when separation was done by ultracentrifugation. Widening at half-height of the peaks of other coarser or finer fractions must be controlled. Indeed, if the crystallites are too fine, the reflections widen; if they are too big, absorption phenomena can disturb the intensity. It should be noted that smectites are concentrated in the finest fractions; kaolinites, illites, chlorites, iron oxides in the 0.2–0.5 µm fractions; detrital micas, chlorites, quartz, feldspars in the 0.5–2 µm fractions. A well-crystallized but incompletely delaminated kaolinite can be found in the fine silt phase. Location of the Angular Distances The angular location of the top of the peaks of diffraction must be transformed into interreticular distance “d ” in Å, in accordance with the Bragg equation (1) (see table or software depending on the type of anticathode used). Each peak should numbered chronologically and its reticular distance given in Angströms or nanometers (international standard), along with its relative intensity. Computer equipment allows

X-ray Diffractometry

117

on-screen comparisons with reference spectra and an automatic search of the JCPDS database (note 1 of Sect. 4.2.2). The powder spectrum of the whole soil provides an overview of all the minerals present and of the relative intensities of the different components. The spectrum may not be easy to interpret if this is not done by comparing it with other samples of the same sequence but at least makes it possible to select appropriate samples for a more thorough examination. The clay fractions of the spectrum powders, which are purified and concentrated by dispersion and particle-size separation treatments, provide more precise information on the clays and on associated minerals. It is possible to determine predominant peaks with strong intensity, and assemblies of characteristic peaks at the level of the reticular distances from 7–, 10–, 12–, 14–15 Å, as well as the peaks of quartz, calcite, etc., and possibly the peaks at 4.40–4.50 Å that are identifiable in the majority of clays. The same limitations apply to complex mixtures. The peaks are often broad, and the profiles asymmetrical; smectites only display a reproducible interlayer space if their state of hydration and ionic saturation are controlled. The presence of interstratified minerals can cause problems that are difficult to solve, especially for the di- and tri-octahedral sub-groups because of obstruction of the zone (060) around 1.50Å. Depending on the objectives, a spectrum helps determine the strategy to be used e.g. adjustments of the experimental conditions, or of the number of oriented plates with given treatments, and possibly the need for other complementary techniques such as thermal analysis, IR spectrometry, electronic microscopy, selective dissolutions, chemical purifications, and so on. On oriented spectra that have undergone suitable treatments basal reflections, relative intensities, regrouping of the diagnostic reflections using JCPDS-ICDD interpretation tables (note 1 in Sect. 4.2.2), can be performed rapidly on-screen using suitable software. The table designed by Brindley and Brown (1980) can also be used, it takes into account the most intense lines characteristic of clay and associated minerals in soils, classified in descending order of the values of “d ”. The “Hanawalt Mineral Search Index” arbitrarily divides the field of the reticular distances from 999.99– to 1.00 Å into 40 groups. The first entry corresponds to the line of maximum intensity. The value “d ” of the second line corresponds to the second strongest intensity and determines the sub-group. The entries are then classified within each sub-group by decreasing values of intensity resulting in six additional lines. Since the degree of intensity is difficult to determine, multiple entries make it

118

Mineralogical Analysis

possible to integrate experimental variations: if the diffraction of a mineral includes 2 or 3 high-density peaks, there can be 2 or 3 entries (with intensities ranging from 100 to 75). The record of the location of the eight peaks should be followed by the name of the mineral and the number of complete JCPDS cards concerning this mineral. The behaviour of minerals under different chemical and thermal treatments facilitates the identification of clay minerals (see Tables 4.4 – 4.7).

4.3. Quantitative mineralogical analysis
4.3.1 Interest Quantitative mineralogical analysis is used to identify factors that influence or determine current or previous pedogenesis and physical and chemical properties of soils. This type of analysis is the logical continuation of qualitative analysis but many problems are involved and the precision will be influenced by the chemical and structural complexity of the substrate. For a simple substrate with two main mineralogical species, depending on the methods used, the risk of error may be around 3%, but may reach 5–10% with three species and about 30% for mixtures with n species. The presence of substances with shortrange order further increases the degree of inaccuracy and can even prevent determination with certain methods, if this is the predominant mineral form. The methods use either: – a single instrumental technique, XRD being the most widely used – a group of techniques that make it possible to identify the centesimal composition of the sample more satisfactorily, although with increased complexity and at a higher cost. 4.3.2 Quantitative Mineralogical Analysis by XRD Principle The advantage of these analytical methods is their relative simplicity and especially the fact that a single instrumental technique is involved. X-ray intensities obtained for each component of a mixture are directly correlated with the proportion of the component according to the equation:

Ip =

kpWp

µ

,

(4.2)

X-ray Diffractometry

119

where Wp is the weight of compound P in the sample; Ip is the intensity of the diffraction of the pure compound P; k isp the constant depending on the compound P and the experimental conditions and µ is the attenuation coefficient of mass of the mixture. When the substances are well crystallized and present a definite chemical composition, resulting in intense specific XRD reflections with no superposition, quantification is easy and the degree of precision is acceptable, particularly with a standard giving the same chemical characteristics. Unfortunately, the imperfection of soil minerals (structural order and disorder), the size of crystallites, chemical variability due to substitutions, and preferential orientations result in modifications in the intensity of peaks for the same mineral and even in angular displacement (which makes the selection of reference minerals for the calibration of measurements difficult). In this case the resulting measurements are very precise. When possible, the reference minerals should come from the same geological formation and have undergone the same type of pedological deterioration in order to reproduce a matrix with a similar degree of disorder and chemical composition (clays under transformation processes). The samples for quantitative analysis have to be prepared with particular care to limit widening of the lines (over-grinding can result in an increase in structural disorder as well as in an increase in the amorphous phase), the effects of extinction and micro-absorption (crystallites and particles are too coarse), and phenomena of orientation due to the interrelationships between particles (preferential orientation). Measurements are taken on: – powder samples that account for all the reflections; as the relative intensity of the basal reflections is low, a concentration of approximately 10% may be necessary to quantify a given compound in the mixture; – an oriented sample which increases the basal reflections of clays, however, this involves the risk of certain components obstructing the regularity of the orientation; this technique is more sensitive and pushes back the limits of detection, but instead of preparing the sample on glass slide, the sample should be prepared on porous ceramic to eliminate the effects of sedimentation (cf. Sect. “Aggregates Oriented on Porous Ceramic Plate”). Sample preparation can include Mg2+ saturation, solvation with glycerol and finally, mixing an internal standard and a matrix suppressor. The homogenization of the sample and the internal standard should be

120

Mineralogical Analysis

carried out in a mixer with an agate ball. When clay samples are separated by ultracentrifugation, a check should be made to make sure there are no variations in composition (the relative proportion of each clay in the mixture may be different and cause bias). Three types of quantitative XRD methods can be used: direct analyses without standard, analyses with an external standard, analyses with an internal standard (addition of standard, matrix suppressor, etc.). In direct analysis with no standard the mineral of reference is taken directly in the sample matrix. Interesting relative values will be obtained for the comparison of samples from the same sequence. Analysis with an external standard does not require long preparation, but the choice of standard substances for the calculation of the intensities is difficult as is also the case with the other methods. The degree of precision is thus sometimes doubtful except for compounds that give clear reflections in a relatively pure medium with only few major components, for example well-crystallized kaolinite and quartz. Analysis with an internal standard: these analyses are carried out in the presence of a standard substance presenting (i) a low attenuation coefficient, (ii) a strong XRD reflection that is narrow and is not superimposed on reflections of the minerals to be determined, and, if possible, is low in harmonics and (iii) a density that is close to the minerals in the mixture enabling better homogenization. The αAl2O3 corundum was adopted in 1976 by JCPDS (see note 1 in Sect. 4.2.2) as the standard of reference for the quantitative study of minerals (high-purity synthetic corundum, 1 µm particle size, Linde, by Union Carbide or similar). The basal reflection at 2.085 Å (I = 100 for the 1,1,3 plane) is clear. Crystallinity can be increased by heating at 800°C for 1 h. There is no preferential orientation. The 2.106 Å line of MgO, and the 6.11 Å line of Boehmite γAlOOH can also be used. Several procedures and modes of calculation can be used, e.g. standard additions or matrix suppression. The standard additions method is time consuming and requires the construction of curves of calibration. The technique of Chung (1974 a,b) with matrix suppression is faster and if required, makes it possible to identify the presence of amorphous minerals with short-range organization in the differential balance 4. However it is necessary to obtain spectra whose principal reflections do not include diffuse bands or zones of superposition as the treatment of these signals is too complex. This method enables elimination of the

4

If the soil sample contains undestroyed organic matter or substances that are transparent to X-ray, these substances can be found in the differential balance.

X-ray Diffractometry

121

matrix effects from the intensity – concentration, and all the intensities are obtained with only one scan, which reduces possible instrumental errors. Procedure A spectrum is performed on powder samples, on the pure corundum and on the sample mixture + corundum: – weigh a known weight of corundum (c) and a known weight of sample (A); the proportion of the mixture should be 1:1 – homogenize in a horizontal mixer with a bowl and an agate ball for 20 min – pour the powder into a support and pack slightly; level with a razor blade to obtain a smooth but not oriented surface; the reference plane must be perfect 5. – place on a diffractometer under standard conditions using a Cu tube, a variable slit and a graphite monochromator with a time constant enabling accumulation of at least 20,000 counts per peak (minus background noise) and a 2θ scanning rate of 0.5° min−1 (or even 0.25° 2θ min−1); only scan the zone containing significant peaks – locate the position of the most intense diffracted peaks that are representative of each component, determine their intensity and compare with the internal standard – what does this dash imply? matrix suppressor. Calculations Based on the nature of monochromatic X-ray radiation of a defined wavelength, the nature of the matrix effects (absorption) and the basic equation of Klug and Alexander (1959), Chung mathematically extracted the effects of attenuation of mass: k X Q k X Q Ii = i i i = i i i (4.3) µi X i µt where Ii is the intensity of the X-ray diffracted by a selected plane of component i (unknown);

Certain authors prefer to prepare the samples by pelletization although there is a risk of causing a certain orientation of the powder due to the very strong pressure applied. However, there will be fewer errors connected with surface quality or the density of sites likely to diffract.

5

122

Mineralogical Analysis

ki is the constant which depends on the geometry of the diffractometer and the nature of component i; Xi is the weight of the fraction of component i; Qi is the density of component i; µi is the coefficient of mass absorption (or attenuation coefficient of mass) of the pure component i; µt is the coefficient of absorption of the total sample including component i, the internal standard and possibly a reference material. The last two terms characterize the effect of adsorption which is often difficult to measure with other methods. The introduction of a definite weight of a matrix suppressor (corundum resembling an internal standard) makes it possible to introduce: Xƒ is the weight of the matrix suppressor (ƒ, flushing agent) and Xo is the weight of the sample, and the equation: Xf + Xo = Xf +
n i =1

Xi = 1

n being the number of components of the sample, and ( Ii I f ) ( I 0f Ii0 ) = ( X i Xf ) ( µi µ f ) , where I 0 and I i0 represent the intensities of the X-ray diffracted by a f selected plane of each pure component. By introducing the ratio of intensity of reference Ki = Ii/Ic, and other substitutions, one arrives at the equation: (4.4) X i = Xf ( kf ki ) ( I i I f ) , which gives the relation between intensity and concentration from which the effect of matrix is eliminated and which is used for quantitative multicomponent analysis. With corundum as matrix suppressor the final simple equation (kƒ = k c = 1) is: (4.5) X i = ( Xc ki )( I i I c ) , where Xi is the weight of the sample fraction; Xc is the weight of corundum; Ii is the diffracted intensity of sample; Ic is the diffracted intensity of corundum; ki = Ii/ Ic = intensity ratio of reference (Table 4.8).

X-ray Diffractometry

123

Remarks Using computerized equipment, it is possible to take into account the height of the reflection and the width at mid-height 6 or the surface of the d(001) reflections which the software calculates automatically taking stabilized background noise into account. Complementary measurements may be necessary on oriented samples: Mg2+ saturated or solvated samples, and the use of multiplicative coefficients accounting for the structure of the minerals (fibrous clays with pseudo-layers that do not give very intense reflections, etc.). It is possible to combine the results of the average of two reflections as this can provide information on crystallinity, etc. The choice will depend on the shape of the diffractogram, the nature of the components, and the degree of precision desired.
Table 4.8. Recommended values of Ii /Ic ratios (Eq. 5) from Bayliss (1986) Mineral allophane biotite boehmite calcite corundum dickite gibbsite goethite gypsum hematite illite I/ Ic 0.1 9.0 1.0 3.7 1.0 2.9 1.6 1.4 2.2 0.9 0.7 d (Å) 3.3 10.0 6.11 3.03 2.09 7.2 4.85 4.18 2.87 2.70 10 Mineral illitemontmorillonite kaolinite 1Md kaolinite 1T muscovite quartz smectite talc I/ Ic 0.4 1.1 2.1 2.2 4.3 3.0 1.5 d (Å) 12 7.1 7.1 10 3.35 15 9.3

For the same mineral species, values of k i may vary with the geological origin and the nature of pedological alteration and measurements must must thus be carried out under the same conditions. The choice of minerals is made after chemical analysis and XRD.
6

If it is necessary to compare samples by measuring intensity, it should be noted that the ratio of the heights of the peaks is only valid if the widths at mid - height are identical for the two samples.

124

Mineralogical Analysis

The Ii /Ic ratio is affected by crystallinity, it approaches 0 if the mineral is not crystalline (allophane – materials with short-range organization) and can reach 8–9 if the size of crystallites and crystallinity is optimum. Chemical 2+ substitutions (e.g. heavy Fe minerals replacing light Mg2+minerals) cause variations in the Ii /Ic ratio.
4.3. 3 Multi-Instrumental Quantitative Mineralogical Analysis Quantitative methods based on XRD have limited precision particularly when dealing with complex assemblies that give diffuse reflections or reflections that are more or less masked by superposition, or when there is a significant quantity of substances amorphous to X-rays. Multiinstrumental methods combine measurements based on XRD and other chemical and physical measurements making it possible to characterize the different elements. Measurements are generally made on clay fractions, but these measurements can be supplemented by others, for example organic matter destroyed by hydrogen peroxide, carbonates, soluble salts, iron oxides, etc. eliminated during extraction of the <2 µm clay fraction, and finally by analysis of sands and silts separated by sieving. These methods can only be used in specialized laboratories that have a wide range of instrumental methods such as XRD, IR, TEM, DTA-TGA, AA, ICP, etc., at their disposal. Each component of a mixture has its own chemical and physical characteristics that can be measured. First XRD spectra are qualitatively interpreted to identify clayey minerals (cf. Sect. 4.2.4 and 4.2.5). Roberts (1974) and Robert et al. (1991) quantified the different elements using their specific properties. The organization, size, and shape of the particles make it possible the right choices and to enhance identification by methods like TEM-HR, STEM, EDX (see Chap. 8) of minerals that only present in small quantities and cannot be detected by XRD. Thermal analysis may be essential for the quantification of kaolinite, oxyhydroxides and chlorite. Losses between 110–300, 300– 600, 600–950°C are used. Corrections for the oxidation of iron at high temperatures are required. Total chemical analysis of clay by a HF–HCl attack (cf. Chap. 31) makes it possible to determine the proportions of the different elements present, which is essential for the identification of the structure of the different mineral phases. Total K enables estimation of micas on the basis of 7.5% of K for illites compared to 8.3 – 10% of K for the other micas. Analysis by “selective” dissolution (cf. Chap. 6) using suitable procedures enables certain phases to be preferentially dissolved without

X-ray Diffractometry

125

the other elements in the matter undergoing a significant attack. For example, the attack of a sample (heated at 110°C for 4 h) by a 0.5 mol (NaOH) L−1 solution followed by boiling for 2 min 30 makes it possible to extract allophane (42.7% SiO2–36.3% Al2O3) and noncrystalline compounds such as colloidal silica (a little montmorillonite and vermiculite is dissolved). Fusion with pyrosulphate or possibly a tri-acid attack (H2SO4–HNO3–HCl) enables isolation of quartz and feldspars in the residue from the attack. This residue is then weighed. The weight should be increased by 3% to compensate for the slight dissolution of quartz and feldspars. The residue is then analyzed by XRD to detect the presence of feldspar then can be analyzed chemically after dissolution. Dissolution with Tamm reagent (in darkness or with UV photolysis) enables isolation of noncrystalline forms of iron and the CBD method (see Chap. 6) enables isolation of crystalline iron hydroxides. Analyses that identify the activity of clays can enable separation of certain types of 2:1 clay based on their cation exchange capacity using several different treatments: Na+ saturation and displacement by Mg++, Mg2+ saturation and displacement by ammonium acetate, etc. The total specific surface (external and/or internal) can be determined using the (EGME) method, by absorption of methylene blue, or by the BET method.
Table 4.9. Some properties of minerals used for the adjustment of the results CEC (cmol kg ) mica quartz feldspars allophane kaolinite colloidal SiO2 montmorillonite vermiculite chlorite 25 2 2 100 5 – 10 20 80 – 100 175 – 200 25
−1

Specific surface 2 (m g−1) 175 25 25 800 – 1,000 45 200 800 440 175

Water loss 540–900°C (%) 0.50 0 0 0 0 2 0.88 1.83 12.30

All this quantitative information is imported into the software making it possible to specify the proportion of each component in the mixture, to

126

Mineralogical Analysis

assign the limits of the properties for the various compounds, and to select options for calculations.

References
Anton O and Rouxhet PG (1977) Note on the intercalation of kaolinite, dickitte and halloysite by dimethyl-sulfoxide. Clays Clay Minerals, 25, 259–263 Bayliss P (1986) Quantitative analysis of sedimentary minerals by powder XRay diffraction. Powder diffraction, 1, 37–39 Brindley GW and Brown G (1980) Crystal structure of clay minerals and their X-Ray identification. Mineralo. Soc., 415–438 Calvert CS (1984) Simplified, complete CsCl–hydrazine–dimethylsulfoxide intercalation of kaolinite. Clays Clay Miner., 32, 125-130 Chassin P (1974) Influence de la stéréochimie des diols sur la formation des complexes interfoliaires de la montmorillonite calcique. Clay Miner., 11, 23–30 Chung FH (1974a) Quantitative interpretation of X-Ray diffraction pattern of mixtures. I- matrix flushing method for quantitative multi-component analysis. J. Appl. Crystallog., 7, 519–525 Chung FH (1974b) Quantitative interpretation of X-Ray diffraction patterns of mixtures. II – Adiabatic principle of X-Ray diffraction analysis of mixtures. J. Appl. Crystallog., 7, 526–531 Churchman GJ, Whitton JS, Claridge GGC and Theng BKG (1984) Intercalation method using formamide for differentiating halloysite from kaolinite. Clays Clay Miner., 32, 241–248 Decarreau A (1990) Les poudres : techniques expérimentales et interprétation des diagrammes – Facteurs déterminant le mode d’empilement. In Structure, propriétés et applications. Société Française de Minéralogie et cristallographie, Groupe Français des Argiles, 209–236 Eltantawy IM and Arnold PM (1974) Ethylene glycol sorption by homoionic montmorillonites. J. Soil Sci., 25, 99–110 Gonzalez Garcia S. and Sanchez Camazano M. (1968) Differenciation of kaolinite from chlorite by treatment with dimethylsulfoxyde. Clay Miner. Bull., 7, 447–450 Klug HP and Alexander LE (1974) X-Ray diffraction procedures. Wiley 2nd edition Modre DZ and Dixon JB (1970) Glycerol vapor adsorption on clay minerals and montmorilllonite soil clays. Soil Sci. Soc. Am. Proc., 34, 816–822 Olejnik S, Aylmore LAG, Posner AM and Quirk JP (1968) Infra-red spectra of kaolin mineral-dimethyl sulfoxide complexes. J. Phys. Chem., 72, 241–249 Paterson E, Bunch JL and Duthie DML (1986) Preparation of randomly oriented samples for X.Ray diffractometry. Clay Miner., 21, 101-106 Range KJ, Range A and Weiss A (1969) Fireclay type kaolinite or fire-clay mineral? Experimental classification of kaolinite – halloysite minerals. Proceedings of the International Clays Conference (Tokyo). Israel Universities Press, 3–13

X-ray Diffractometry

127

Roberts JM Jr (1974) X-Ray diffraction and chemical techniques for quantitative soil clay mineral analysis. Engineering Thesis, Pennsylvania State University, 78 pages Robert M, Hardy M and Elsass F (1991) Crystallochemistry, properties and organization of soil clays derived from major sedimentary rocks in France. Clay Miner., 26: 409–420 Thomson A, Duthie DM, Wilson MT (1972) Randomly oriented powders for quantitative determination of clay minerals. Clay Miner., 9, 345–348 Wada K and Yamada H (1968) Hydrazine intercalation, intercalation for differentiation of kaolin mineral, from chlorites. Am. Miner., 53, 334–339

Bibliography

General
Alekseeva TV, Alekseev AO, Sokolovska Z, Khainos M, Sokolowska Z and Hajnos M (1999) Relationship between mineralogical composition and physical properties of soils. Pochvovedenie., 5, 604–613 Brindley GW and Brown E (1980) Crystal structure of clay minerals and their X-Ray identification., Mineralogical Society, 495 p Caillère S, Hénin S and Rautureau M (1982) Minéralogie des argiles, 1, Masson, 184 p Caillère S, Hénin S and Rautureau M (1982) Minéralogie des argiles, 2, Masson, 189 p Charley H (1989) Clay Sedimentology., Springer Belin Heidelberg New York, 623 p Dixon JB and Weed SB (1989) Minerals in soil environments., Soil Science Society of America (USA), 2e édition, 1244 pp. Gautheyrou J and Gautheyrou M (1979) Etude des argiles par diffraction X. Synthèse bibliographique pour l'identification des argiles., Guide pratique. ORSTOM-Antilles, notes de laboratoires, ORSTOM (26 pages + 2 annexes). ICDD (JCPDS-ASTM)) Mineral powder diffraction File – PDF-1 DATA BASE (powder diagramms, interlayer spaces, relative intensity, chemical name, mineralogical formula) - PDF-2 DATA BASE (powder diagramm, interlayer spaces, relative intensity, chemical name, mineralogical name, Miller’s indice, unit cell, physical properties, references) – Shorten ICDD ref: eliminate – version papier (SET 1 à 36, SET 1 à 8 révisés, SETS 37, 38, 39, 40, 41) – version disque compact (CD-ROM DISC SETS 1-41 inorganique – organique pour IBM PC, VAX, McIntosh) – SEARCH MANUAL Alphabetical index – inorganic phases. Hanawalt index – inorganic phases (remise à jour annuelle), ICDD Newton Square PA 19073-3273 (USA) Inigo AC, Tessier D, Pernes M (2000) Use of X-ray transmission diffractometry for the study of clay-particle orientation at different water contents. Clays Clay Miner., 48, 682–692

128

Mineralogical Analysis

Kovda IV, Morgun EG, Tessier D, Pernes M (2000) Particle orientation in clayey soils according to transmission diffractometry data, 8, 989–1003 Manhães RST, Auler LT, Sthel MS, Alexandre J, Massunaga MSO, Carrió JG, dos Santos DR, da Silva EC, Garcia-Quiroz A and Vargas H (2002) Soil characterisation using X-ray diffraction, photoacoustic spectroscopy and electron paramagnetic resonance. Appl. Clay Sci., 21, 303–311. Martins E de S, de S Martins E (2000) Integrated method of mineralogical characterization of deeply weathered soils. Comunicado Tecnico Embrapa Cerrados., Brazil, 37, 5 pp. Millot G (1964) Géologie des argiles. Masson, Paris, 499 pp. Newman AC (1987) Chemistry of clays and clay minerals., Mineralogical society, monograph no 6, 480 p Chen PY (1977) Table of key lines in X-Ray powder diffraction patterns of minerals in clays and associated rocks. Dept. of Natural resources, (Indiana, USA). Geological Survey occasional paper no 21, 67 p Robert M (1975) Principes de détermination quantitative des minéraux argileux à l'aide des rayons X. Problèmes particuliers posés pour les minéraux argileux les plus fréquents dans les sols des régions tempérées. Ann. Agron., 26, 363–399 Stucki JW (Goodman BA and Schwertmann U (1985) Iron in soils and clay minerals., D. Reidel, 893 p. Teissier D (1984) Etude expérimentale de l’organisation des matériaux argileux. Hydratation, gonflement et structuration au cours de la dessiccation et de la réhumectation., INRA, Thèse doc. Etat, 361 pp. Thorez (1975) Phyllosilicates and clay minerals. A laboratory handbook for their X-Ray diffraction analysis., Lelotte Ed., 579 p Wilson MJ (1987) A handbook of determinative methods in clay mineralogy., Blackie – Chapman and Hall, 308 p.

Preparation of Oriented Aggregates on Porous Ceramic Plate
Kinter EB and Diamond S (1956) A new method for preparation and treatment of oriented-aggregat specimens of soil clays for X-Ray diffraction analysis. Soil Sci., 81, 111–120 La Manna JM and Bowers FH (1985) A suction apparatus for orienting clay minerals into porous ceramic tile. Soil Sci. Soc. Am. J., 49, 1318–1319 Rich CI (1969) Suction apparatus for mounting clay specimens on ceramic tile for X-Ray diffraction. Soil Sci. Soc. Am. Proc., 33, 815–816 Shaw HF (1972) The preparation of oriented clay mineral specimens for X-Ray diffraction analysis by a suction unto ceramic tile method. Clay Miner., 9, 349–350

X-ray Diffractometry

129

Saturation of Clays by Cations
Brindley GW and Ertem G (1971) Preparation and solvation properties of some variable charge montmorillonite. Clays Clay Miner., 19, 399–404 Bühmann C, Fey MV and De Villiers JM (1985) Aspects of the X-Ray identification of swelling clay minerals in soils and sediments. S. Afri. J. Sci., 81, 505–509 Calvet R and Prost R (1971) Cation migration into empty octahedral sites and surface properties of clays. Clays Clay Miner., 19, 175–186 Hofmann U and Klemen R (1980) Verlust der Austanschfahigeit von lithiumionen an bentonit darch erhitzung (perte d’échangeabilité des ions lithium dans les bentonites après chauffage). Zeit. Anorganisc Chem., 262, 95–99 Lim CH and Jackson ML (1986) Expandable phyllosilicate reactions with lithium on heating. Clays Clay Miner., 34, 346–352

Saturation, Solvation, Intercalation Complexe, Dissolution
Barnhisel RI and Bertsch PM (1989) Chlorites and hydroxy-interlayered vermiculite and smectite. In Minerals in soil environments Dixon JB and Weed SB ed., Soil Sci. Soc. of Am., 729–740 Barnhisel RI (1977) Chlorites and hydroxy interlayered vermiculite and smectite, 331-356. In Minerals in soil environments, Dixon JB and Weed SB ed., Soil Sci. Soc. Am., Monogr., 331–356 Brindley GW (1966) Ethylene glycol and glycerol complexes of smectites and vermiculites. Clay Miner., 6, 237–260 Brindley GW and Ertem G (1971) Preparation and solvation properties of some variable charge montmorillonites. Clays clay Miner., 19, 399–404 Churchman GJ (1990) Relevance of different intercalation tests for distinguishing halloysite from kaolinite in soils. Clays and clay Minerals, 38, 591–599 Follet EAC, McHardy WJ, Mitchell BD and Smith BFL (1965) Chemical dissolution techniques in the study of clays. Part 1. Clay Miner., 6, 23–24 Novich BE and Martin RT (1983) Solvation methods for expandable layers. Clays Clay Miner., 31, 235–238 Suquet H, Iiyama JT, Kodama H and Pezerat N (1977) Synthesis and swelling properties of saponites with increasing layer charge. Clay Miner., 25, 231– 242 Suquet M, Calle de la C and Pezerat H (1975) Swelling and structural organization of saponite. Clays Clay Miner., 23, 1–9 Theng BKG, Churchman GJ, Whitton JS and Claridge CGC (1984) Comparison of intercalation methods for differentiating halloysite from kaolinite. Clays Clay miner., 32, 249–258 Walker GF (1958) Reactions of expanding-lattice clay minerals with glycerol and ethylene glycol. Clay Miner. Bull., 302–313

130

Mineralogical Analysis

White JL and Jackson ML (1947) Glycerol solvation of soil clays for X-Ray diffraction analysis. Soil Sci. Soc. Am. Proc., 11, 150–154

Preparation of Iron Oxides
Brown G and Wood IG (1985) Estimation of iron oxides in soil clays by profile refinement combined with differential X-Ray diffraction. Clay Minerals, 20, 15–27 Campbell AS and Schwertmann U (1985) Evaluation of selected dissolution extractants in soil chemistry and mineralogy by differential X-Ray diffraction. Clay Miner., 20, 515–519 Meunier A and Velde B (1982) X-Ray difffraction of oriented clays in small quantities (0,1 mg). Clay Miner., 17, 259–262 Paterson E, Bunch SL and Duthie DML (1986) Preparation of randomly oriented samples for X-Ray diffractometry. Clay Miner., 21, 101–106 Schwertmann U and Taylor RM (1989) Iron oxides. In Minerals in Soil environments, Dixon JB and Weed SB ed. Soil Sci. Soc. Am., 379–438 Schwertmann U, Murad E and Schulze DG (1982) Is there Holocene reddening (hematite formation) in soils of a xeric temperate area. Geoderma, 27, 209– 223 Torrent J, Schwertmann U and Schulze DG (1980) Iron oxyde mineralogy of some soils of two river terrace sequences in Spain. Geoderma, 23, 191–208

Quantitative XRD
Austin GS and Leininger RK (1976) Effect of heat-treating mixed-layer illitesmectite as related to quantitative clay mineral determinations. J. Sedim. Petrol., 46, 206–215. Brime C (1985) The accuracy of X-Ray diffraction methods for determining mineral mixtures. Miner. Mag., 49, 531–538 Carter RJ, Hatcher MT and Di Carlo L (1987) Quantitative analysis of quartz and cristobalite in bentonite clay based products by X-ray diffraction. Anal. Chem., 59, 513-519 Cody RD and Thomson GL (1976) Quantitative X-ray powder diffraction analysis of clays using an oriented internal standard and pressed discs of bulk shale. Clays Clay Miners, 24, 224–231 Davis BL (1980) Standardless X-Ray diffraction quantitative analysis. Atmosph. Environ., 14, 217–220 Decleer J (1985) Comparaison between mounting techniques for clay minerals as a function of quantitative estimations by X-ray diffraction. Bull. Soc. Belge Géol., 94, 275–281 Gavish E and Friedman GF (1973) Quantitative analysis of calcite and Mg– calcite by X-ray diffraction effect of grinding on peak height and peak area. Sediment., 20, 437–444

X-ray Diffractometry

131

Goehner RP (1982) X-Ray diffraction quantitative analysis using intensity ratios and external-standards. Adv. in X-Ray Analy., 25, 309–313 Heath GR and Pissas NG (1979) A method for the quantitative estimation of clay minerals in North Pacific deep sea sediments. Clays Clay Miner., 27, 175–184 Hogson M. and Dudgney ANL (1984) Estimation of clay proportions in mixtures by X-Ray. Diffraction and computerized chemical mass balance. Clays and Clay Miner., 32, 19–28 Hooton DH and Giorgetta NE (1977) Quantitative X-Ray diffraction analysis by a direct calculation method. X-Ray Spectrom., 6, 2–5 Hubbard CR and Smith DK (1977) Experimental and calculated standards for quantitative analysis by powder diffraction. Adv. X-Ray Anal., 20, 27–39 Hubbard CR, Evans EH and Smith DK (1976) The reference intensity ratio I/Ic for computer simulated powder patterns. J. Appl. Cryst., 9, 169–174 Johnson LJ, Chu CH and Hussey GA (1985) Quantitative clay mineral analysis simultaneous linear equations. Clays and Clay Miner., 33, 107–117. Kahle M, Kleber M and Reinhold J (2002) Review of XRD-based quantitative analyses of clay minerals in soils: the suitability of mineral intensity factors. Geoderma, 109, 191–205 Norrish K and Taylor RM (1962) Quantitative analysis by X-Ray diffraction. Clay Miner. Bull., 5: 98–109 Ouhadi VR and Yong RN (2003) Impact of clay microstructure and mass absorption coefficient on the quantitative mineral identification by XRD analysis. Appl. Clay Sci., 23, 141–148 Parrot JF, Verdoni PA and Delaune-Mayere (1985) Analyse modale semiquantitative d’après l’étude des Rayons X. Analusis, 13, 373–378 Pawloski GA (1985) Quantitative determination of mineral content of geological samples by X-Ray diffraction. Am. Mineral., 70, 663–667 Persoz (1969) Fidélité de l’analyse quantitative des poudres de roches par diffration X. Bull. Centre Rech. Pau (SNPA), 3, 324–331 Renault J (1987) Quantitative phase analysis by linear regression of chemistry on X-Ray diffraction intensity. Powder Diffract., 2, 96–98 Ruffell A. and Wiltshire P. (2004) Conjunctive use of quantitative and qualitative X-ray diffraction analysis of soils and rocks for forensic analysis. Forensic Science International, 145, 13–23 Taylor RM and Norrish K (1972) The measurement of orientation distribution and its application to quantitative determination of clay minerals. Clay Miner., 9, 345–348 Tomita K and Takahashi H (1985) Curves for the quantification of mica/smectite and chlorite/smectite interstratifications by X-Ray powder diffraction. Clays Clay Miner., 33, 379–390.

5

Mineralogical Analysis by Infra-Red Spectrometry

5.1 Introduction

5.1.1 Principle The interaction of matter with infra-red (IR) radiation makes it possible to characterize energies of vibration of the molecules on several components (Fig. 5.1): along the axis of the chemical bonds (vibrations of valence or stretching, which, apart from diatomic molecules, are seldom pure) and deformations that are perpendicular to the bond axis (rotation, torsion, shearing, swinging, librations, bending). IR radiations correspond to these energy levels. IR absorption occurs when the frequency of the radiation is equal to that of the vibrations.

Fig. 5.1. Examples of vibration of a simple polyatomic molecule

134

Mineralogical Analysis

IR adsorption spectroscopy uses radiations ranging between the visible waves and microwaves. This field is usually divided into three energy zones: near-IR, medium IR and far IR (Fig. 5.2). In these zones, different molecular vibrations correspond to the energy of IR radiations. – Near infra-red (NIR), as well as visible and UV, account for the highenergy electronic spectra related to fundamental orbitals, for example the change from a link orbital to an empty orbital of higher energy; NIR spectrometry is currently being developed for the study of soil organic matter (cf. Sect. 5.3.1 and Part 2 of this book). – Medium IR, ranging between 300 and 5,000 cm–1 makes it possible to observe vibrations involving protons, (vibrations that correlate well with structure and whose transitions correspond to slight modifications in stretching or deformation of the bond angles in the molecule). Unit cells of clays (crystallographic units) contain polyatomic ions or molecules whose internal modes of vibrations occur between 4,000 and approximately 400 cm–1. These vibratory states have been the subject of detailed studies in mineralogy. Absorption bands make it possible to characterize active molecular groupings satisfactorily. – Other modes of vibrations which come from the lattice can occur after displacement of a polyatomic group within a unit cell in far IR at very low frequencies between 200 and 10 cm–1. This field, which has not been extensively explored to date, is now accessible thanks to the development of IR spectrometers. Bands of rotational transitions that are not widely spaced enable quantification of the number of revolutions around an axis without stretching or notable modification of the bound angles that are characteristic of the geometry of the crystal. IR spectrometry is thus used as a complement for X-ray diffraction and chemical and thermal analysis. XRD (cf. Chap. 4) expresses longdistance periodicity satisfactorily, but is not effective in the case of substances that are amorphous to X-ray or minerals with short-range arrangement that appear during sequences of alteration. With IR, a spectral signature can be identified, along with the nature and the direction of the bonds, and our understanding improved of atomic structures, as well as of the degree of isomorphic substitution in the tetrahedral (Si–Al) and octahedral (Al–Mg) layers. These data are needed to identify certain minerals, to quantify molecular water and constitutive hydroxyls and to detect the presence of crystalline or non-crystalline impurities, which influence the regularity of the lattice structure. In clays and clay minerals only some molecular groups are likely to vibrate and the spectra are often less complex than for certain organic substances.

IR Spectrometry

135

l n n

Fig. 5.2. Field of infra-red molecular spectrometry and transparency of the optical elements. The position of the bands in x-coordinate is expressed either in wavelength λ in nm or µm, or often in wavenumber 1 104 . In ordinate, transmittance expresses percent ν (cm−1 ) = = λ (cm) λ ( µ m ) of radiation that crossed the sample (PE = polyethylene, Mylar)

5.1.2 IR Instrumentation A spectrometer includes an IR source, optical elements, a detector and a rack of computerized measurements. It is difficult to choose optical equipment that covers the whole spectrum from near IR to far IR (Fig. 5.2). The source must emit intense polychromatic radiation covering the whole of the IR spectrum. A filter eliminates UV and visible radiations that are emitted simultaneously. – Sources with filaments of tungsten cover only the field of near IR; – nickel-chromium filaments make it possible to reach 600 cm–1; – mercury vapour lamps make it possible to reach far IR between 300 and 10 cm–1; – globars (carborundum rods with refractory oxides), which are often used with water cooling to stabilize the temperature of the source at around 1,500°C, emit up to approximately 200 cm–1; – silicon carbide sources emit between approximately 6,500 and 50 cm–1. Optical equipment (e.g. lenses, windows, dispersive systems) should be selected for their transmittance properties (Fig. 5.2) and resistance to water or solvents. The lenses should preferably be replaced by mirrors (Figs. 5.3 and 5.8). Filtering of the visible spectrum is generally accomplished by surface treatments with germanium, or black polyethylene films of varying thickness. The choice of the detector is also limiting and its surface area, the spectral field covered, the sensitivity and the response time of the apparatus, the frequencies to be detected, maintenance needs (operation

136

Mineralogical Analysis

under liquid nitrogen or helium) should all be taken into account. The detector can be: – non-selective e.g. a thermocouple or thermopile (several couples connected in series), a bolometer with doped germanium for far IR that operates in liquid helium (2 K), MCT detectors (Mercury, Cadmium, Tellurium), DTGS detectors (Deuterium enriched Triglycine Sulphates) thermostated for medium IR, detectors of Golay with a gas chamber for far IR; these detectors are very sensitive, but very fragile; – selective (photon–electron transformation as a function of the wavelength), e.g. lead sulphide, lead selenide, or indium antimonide detectors that function in liquid nitrogen. The optical system can be based on a dispersive mode or managed by interferometry in a mono or double beam system. In dispersive mode radiation crosses the sample from where it partially arises (transmission-absorption) to strike a dispersive lattice or a monochromator which divides the beam as a function of wavelength (Fig. 5.3a). Energy is recorded point by point by rotation of the lattice. It is first necessary to determine the zero point of the instrument, and then to determine the basic spectrum without a sample to take into account in particular the CO2 and H2O in the air. Transmittance is calculated at each wavelength by the ratio of the two signal-to-noise values. The resolution is often insufficient and energy decreases with an increase in wavelength, which makes it necessary to gradually open the slit of the monochromator and modify the background noise. These operations can be automated, and the degree of precision can be increased by the use of a double beam and measurements on sample and blank taken alternately at each wavelength with the help of a modulator. In this way the energy of the beam is maintained constant. The study of far IR is not possible because of the nature of the apparatuses and insufficient energy. Ratiometric apparatuses are slow and top-of-the-range models are very expensive, as several high resolution lattices are necessary for satisfactory linear dispersion. Better performances can be obtained by using an interferometric system linked with data processing by Fourier transform (Fig. 5.3b). Interferometry is based on rapid movements of a mirror. Each wavelength is modulated at a characteristic frequency determined by the speed of the mirror. The recording of the complex data gives an interferogram that is treated in real time by the Fourier transform. This makes it possible to obtain a spectrum where the amplitude of the signal is recorded as a function of the frequency.

IR Spectrometry

137

a

b

Fig. 5.3. Types of Infra-red spectrometers: (a) Dispersive 1: dispersive system: prisms, lattices, monochromator, continuous high resolution interferential wedge, 2: exit slit for precise selection of the field of frequency, 3, 4, 5: detection, amplification and acquisition, I0 I, IX : total incidental and transmitted intensities at the selected wavelength x, T: transmittance I/I 0 corresponding to the absorbance A = – log T (b) With interferometer and Fourier transform 1: separation made of KBr with germanium film (400– 4,000 cm–1) or made of MYLAR (0–700 cm–1); 2: mobile mirror, spectral composition is a function of the position of the mirror, function resolution of the amplitude of displacement, 3: He–Ne laser for measuring movements of the mobile mirror and indexing mirror direction, 4: sample compartment under vacuum or controlled atmosphere, 5: intensity/position of mirror, 6: interferogram, 7, 8: real-time data acquisition system

In the Michelson interferometer, the polychromatic radiation of the source is divided into two beams by a separation made of KBr or Mylar (covered with a germanium film to filter radiations of the visible spectrum). One of the two beams is sent to a fixed mirror, the other to a mobile mirror that is moved at a known speed by to a linear motor. The two beams are than recombined by the beam splitter. The resolution depends on the maximum stroke of the mobile mirror. Alignment must be maintained during the movement of the mobile mirror with no vibrations or slope likely to deform the spectra and to give erroneous transmittance values. An in–out movement of the mirror allows simultaneous analysis of all the wavelengths of IR radiation. The changes in the signal reflect variations in modulation of the interferogram. The results are recorded at the different positions of the mirror, which requires knowing its exact position by laser radiation. Each individual signal is located by multiplexing and decoding. The whole spectrum is scanned sequentially as a function of time in less than one second.

138

Mineralogical Analysis

To limit the influence of moisture and atmospheric CO2, the optical circuits function in a sealed enclosure, except for the sample compartment which can be purged with dry nitrogen. The resolution is generally about 2–4 cm–1 but can reach 0.1 cm–1 at the price of additional memory which otherwise requires a powerful information processing system. FTIR apparatuses are generally mono-beam and more rarely, in top-ofthe-range models, double-beam which enables the signal-to-noise ratio to be maintained uninterrupted. Suitable software makes it possible to control the spectrometer, collect the results and process the data in real time to restore the spectrum. If necessary, a database search makes it possible to compare the spectra with IR bases for identification, calibration, subtraction of spectra, etc. Calibration is carried out with polystyrene blades that have narrow intense bands distributed between 699 and 3104 cm–1. In practice, the choice of an apparatus is based primarily on the scientific objectives which determine the main requirements: the wavelength field to be scanned, the nature of materials to be analysed (solid, liquid, gas), the required sensitivity, resolution, data-processing capacity and quality of the software, possible extensions of the basic module (coupled with e.g. liquid or gas chromatography, EGA thermal analysis, Raman spectroscopy, IR or Raman microscopy). As far as price is concerned, instead of purchasing a very expensive top-of-the-range “universal” apparatus, two dedicated complementary apparatuses can be adapted for repetitive laboratory tasks. For routine soil analysis, an apparatus that covers medium IR up to 220–250 cm–1 with an optical system in KBr or better still in CsI and a resolution of about 2 cm–1 is sufficient, but for a more specialized laboratory, access to the field of far IR is now necessary.

5.2 IR spectrometry in Mineralogy
5.2.1 Equipment and Products – Ultra-microbalance, 10–6 g sensitivity, range 2–5 g – IR or FTIR spectrometer suitable for the required ranges of wavelengths

IR Spectrometry

139

– IR microscope – binocular magnifier (×100) – pellet press (10 tons cm–2) – centrifuge with centrifugation chambers (cf. Chap. 4) – cooled moisture-proof vibrating grinder with agate ball for crushing in liquid medium – drying oven (105°C) with precise electronic regulation – Pyrex glass desiccator – lab glassware – quality IR products: KBr, CsI, polyethylene, Nujol (paraffin oil), AgCl, CaF2, IR TRAN 4, solvents, hexachlorobutadiene (Voltalef oil), Fluorolube (fluorinated hydrocarbon oil), various mineral standards.

5.2.2 Preparation of the Samples Types of Preparation The preparation of the sample for analysis by IR spectrometry is of prime importance: it conditions the spectral field of analysis and its limits, and indirectly, the sensitivity and the selectivity of measurements. Determination must be performed on elementary particles that do not exceed 5 µm and even 1 or 2 µm if the MULL technique is used or if quantitative analyses are required. Clay fractions with a particle size <2 µm in H+, NH4+ or X+ form, or better fractions <0.5 µm purified with a standard Sharples ultra-centrifuge (cf. Chap. 3) are usually used for analysis. Because of the size of their unit cell, clays display only one negligible distortion of the absorption bands. Crush the samples and homogenize with the agate mortar in the presence of a volatile organic liquid that is inert to IR (Ethanol, acetone, etc.); avoid modifying or destroying the crystalline structure. Too large particles could cause spectral distortion, dispersion of incidental radiation and widening of the absorption bands (Christiansen effect). Destruction of the organic matter is often necessary first to limit organic absorbance, which is likely to overload the spectrum, and second to avoid excessive retention of adsorbed water, which can mask some absorbance of minerals. However, in some cases this treatment can lead to neogenesis of minerals.

140

Mineralogical Analysis

For measurements on solid samples – In transmission–absorption, clays can be analysed in three main forms: a. thin films that are self-supporting or placed on suitable supports, b. discs made with binding agents that are transparent to IR, c. in a mixture with liquid mulls. – In NIR or Raman diffuse reflectance, the sample is simply packed in a sample holder taking care to limit preferential orientations. – In multiple specular reflectance or attenuated total reflectance (ATR), the problems of interface with the soils are often not reproducible and these techniques can generally only be used on thin blades. – In IR microscopy samples can be analysed on film or as microsamples without preparation. Pretreatments can cause changes in the properties of the samples: these can be used in comparisons with chemically untreated rough samples preserved in their original conditions (amorphous phases, etc.). For measurements on liquid samples separated from the soil by chemical means, it is possible to perform the measurements either (i) in solvents that are transparent to IR and do not react with the minerals, or (ii) NIRS measurements on freeze-dried extracts. For measurements on gas samples resulting from controlled pyrolysis or decomposition of mineral fractions during thermal analyses (EGA and TGA-DTA, see Chap. 7), sample “powders” are used that are identical to those used for standard thermal analyses. To improve sensitivity, special cells are used to lengthen the path of the beam in the gaseous medium, absorption being proportional to length. Preliminary purging of the air in the sample compartment is required to eliminate any H2O and CO2 present. Preparation of Self-Supporting Films Technique This preparation is only possible with certain types of clay, e.g. montmorillonites, vermiculites and fibrous clays, that can be prepared as thin but stable films. The method of Farmer and Palmieri (1975) has been slightly modified to allow quantification of the measurements. This technique has the advantage of not subjecting the sample to a strong pressure and of avoiding exchange reactions between the sample and the binding agent added in the pelletizing method. The main difficulties are the critical thickness of the film (maximum 4–8 µm), its capacity to transmit a sufficient intensity, its mechanical resistance and the frequent difficultly

IR Spectrometry

141

involved in its removal, which requires great technical skill. Silicated minerals often deposit with preferential orientation, which makes it possible to identify the vibrations caused by the oscillation of the dipoles which should be perpendicular to the plan of the lattice. It is also possible to study polychroism related to the plane using a goniometer.

a

b

Fig. 5.4. Preparation of self-supporting mineral films: (a) centrifugation at 2,500 g in Cyto Hettich chamber, (b) removal of self-supporting film

Procedure – Crush a clayey extract of known weight with an agate mortar with a little water to produce a fluid paste, and then put in suspension in a known volume of distilled water – transfer the complete suspension on a slide covered with a flexible polythene film in a Cyto Hettich chamber (Fig. 5.4); centrifuge at moderate speed (approximately 2,500 g) the selected volume of the chamber is 4 mL, making it possible to obtain a film 12.4 mm in diameter, that is to say a surface area of 120 mm2; in this way it is possible to determine the quantity deposited per unit area (the density must be approximately 1–2 mg per cm2) – remove the clear supernatant solution with a pipette and air dry the film containing the deposit – remove the film by passing the flexible support across the edge of a bevel-edged blade and transfer it onto a support for IR measurement; the deposit can also be transferred onto a thin filter support with rigid

142

Mineralogical Analysis

polymeric mesh, which avoids migration thanks to its continuous structure - store the deposit for 48 h in a desiccator on P2O5 before analysis. Preparation of Film on a Support Transparent to IR Principle This system makes it possible to obtain thin films either for oriented deposit by simple gravity or by centrifugation as above. The following factors have to be taken into account: the thickness of the clay deposit, the choice of a suitable solvent to avoid dissolution of the support, the spectral field useable with this type of support, the reactivity of the clay–solvent-support. In general, KBr slides are used because they are cheap and easy to use, and enable medium-IR scanning up to approximately 400 cm–1, or CsI up to 220 cm–1. Polythene is used for far IR. The sample is put in suspension in a solvent without dissolution. The surface can be impregnated by microvaporization with Nujol to limit reflectance phenomena at the air–clay interface. Procedure Depending on the IR domain, select a pellet 13 mm thick obtained by pressure (10 tons cm–2 AgCl, CaF2 IRTRAN 2, IRTRAN 4, Ge, Si, KBr, CsI, polythene, etc.). The sample in suspension in a suitable organic solvent is deposited in the same way as in the preceding procedure by gravity or centrifugation. After drying, carefully heat the disc covered with film in a drying oven at 100°C for 5 h to eliminate all traces of water, then store in a desiccator until analysis. Preparation of Discs (Solid Solution) Principle The solid sample is completely pulverized with an agate mortar in the presence of an organic liquid (for example ethanol) then dried under vacuum in the desiccator with phosphoric anhydride. It is then mixed with a matrix that enables to form self-sustaining discs at high pressure; these disks can be used with quite a wide range of IR radiation. For swelling smectites, it is preferable to grind a known weight of sample with a little water to form a thick paste, then add the binding agent and grind the wet sample again. After complete drying, homogenize the diluted sample with a microball grinder. This system is easy to use and is the most widely used, but it has certain disadvantages:

IR Spectrometry

143

– all materials used as binding agents for the discs have limited transmission in IR so it is impossible to observe the absorbance of a clay on the whole range from near IR to far IR on only one pellet – the reactivity of the support may result in exchange reactions with clay. For example potassium bromide (KBr), which is transparent up to 400 cm–1 (Fig. 5.2), produces excellent pellets that allow good spectra in a wide range of IR. But with 2:1 clays, such as smectites which contract with K+, exchange phenomena can cause deformation of the absorption spectra (Nyquist and Kagel, 1971). These discs are thus not suitable for studies concerning adsorbed water as K reduces water retention, or for the study of surface cations. As KBr is hygroscopic, the pellets have to be stored in a desiccator under vacuum with phosphoric anhydride to limit the phenomena of adsorption of water and resulting uncertainties in interpretation between the bands of hydroxyls and those of water adsorbed by minerals and the support. For far IR, polyethylene, polytetrafluoroethylene (Teflon), or paraffin should be used. Pressure can cause inappropriate transformations by amplifying the effect of the chemical reactivity, but can be used for in situ study of the effects of very high pressures obtained with of a diamond-cell anvil making it possible to analyse by IR the induced transformations (Weir et al., 1959; Liu and Mernagh, 1992). Procedure (Qualitative Analysis) – Choose the matrix and the diameter of the pellets – dry the sample in a desiccator for 48 h to eliminate non-structural water whose 3,440 cm–1 band can mask that of structural OH; this method of drying is not suitable for all minerals (cf. Chap. 1) – dry the finely crushed IR-quality binding agent in the drying oven under vacuum at 100°C for one night if its thermal stability permits – for a disc of approximately 1 mm thickness and a diameter of 13 mm, weigh 0.5–3 mg of clay sample at 0.01 mg precision. – add 300 mg of KBr (can be changed) – homogenize with a microhomogenizer with a plastic or agate ball for 2 min, or grind with the agate mortar once more for a perfect mix – transfer in a stainless steel mould 13 mm in diameter (A in Fig. 5.5b). – apply light pressure and degas under vacuum for 5 min

144

Mineralogical Analysis

a

b

Fig. 5.5. Preparation of the samples in solid solution (a):12 ton manual hydraulic press, (b): detail of a pelletizer: A body, B: removable base, C 13 mm φ : : plunger, D: 13mm φ polished cylinders, E: release ring from the mould, filled circle represents sealing ring.

– Press (Fig. 5.5a) at 10 tons cm–2 for 10 minutes (KBr becomes plastic at this pressure) – extract the disk from the mould using the release ring (E in Fig. 5.5); it should look homogeneous, smooth and transparent; do not touch it with the hands – desiccate at 100°C for 2 h and store in a desiccator on P2O5 until needed. In practice, it is advantageous to make two of even three pellets using the same binding agent at two different concentrations: – a pellet made with 3 mg of sample for 300 mg of KBr to reach total absorption in the zones around 1,000 cm–1 and 500 cm–1 (silicates) – a pellet made with 0.25–0.5 mg of clay sample to obtain details of the spectrum in the areas of intense absorption of silicates – a third pellet made with a binding agent transparent to far IR will also be needed if spectra below 200 cm–1 are required – perform the spectra under the instrumental conditions selected. The time of passage in the spectrometer will depend on the type of material (dispersive or interferometer), the scanning zone with dispersive apparatuses, or the resolution required. After measurement, the pellets can be stored in the desiccator on P2O5.

IR Spectrometry

145

To avoid corrosion, the pelletizer must be cleaned immediately after use without abrasion. Preparation of the Clay Samples in the Form of Mull Principle When interactions are possible between the binding agent and the sample, when it is impossible to make self-supporting films or when pressure is not desirable, the Mull technique can be used with a non-volatile inert oil. Nujol (paraffin oil), hexachlorobutadiene (Voltalef oil) or Fluorolube (fluorinated hydrocarbon oil) is mixed with the sample to form a fluid paste which is pressed between two windows. The method is rapid but only qualitative; the sample is oriented. It is not possible to desiccate the mixture in the drying oven. Procedure – Place 10 mg of dried clay sample in a 50 mm agate mortar; add a known quantity of Mull to moisten the powder using a micropipette or a spatula – crush to obtain a thick paste in which the sample is uniformly dispersed (concentration will be about 0.3–0.5%) – spread out the mixture with a spatula on a window transparent to IR, then cover with another window to obtain a regular thickness taking care not to trap air inside – place in the spectrometer and record the spectrum. The spectral limits of the windows and the zones of absorption specific to the Mull matrix can be taken into account in interpretation. For example, Nujol strongly absorbs between 3,000 and 2,800 cm–1 (CH), 1,460 cm–1, 1,375 cm–1, Fluorolube does not absorb between 4,000 and 1,400 cm–1. Both mulls are complementary and two mulls should be made to check that the absorption bands do not mask the bands specific to the sample. The homogeneity of the suspension is difficult to obtain and maintain: prepared samples should thus always be stored horizontally. Preparation for Specular or Diffuse Reflectance, Attenuated Total Reflectance (ATR) As the intensity of absorption depends on the angle of incidence, the surface should present weak granulation, and isogranulometric grinding to 0.2 mm is thus necessary. Compression should be carried out as for XRD powders (cf. Chap. 4) avoiding excessive orientation.

146

Mineralogical Analysis

The thickness of the powder (approximately 1 mm) is not critical, since the radiation only penetrates a few microns. In certain cases, it is possible to use compressed pellets with or without the addition of binding agents, but in this case the orientation is rather strong. It should be noted that since the refraction index is significant in measurements by reflectance, significant differences will be observed between the spectra obtained by transmission and by reflectance at high wavelengths. Deuterization In specialized laboratories, deuterization is an ideal method to study water in clays. In heavy water, deuterium replaces hydrogen. When H2O is replaced by D2O, the OH of the interstitial water is deuterated, but not the OH of the lattice (Wada, 1966). The interatomic distances do not change, the mass is doubled and the vibration frequencies drop. It is thus possible to separate the reticular OH or adsorbed water, and eliminate the ambiguity of the measurements in studies of mineral gels (Nail et al., 1976). Remarks During preparation, atmospheric contamination of the rough samples must be avoided, for example: – ammonium can produce absorption around 3,250 and 1,400 cm–1 (stretching and deformation) – in the presence of calcium, attacking organic matter with hydrogen peroxide can lead to the formation of insoluble calcium oxalate which produces absorption near 1,400 cm–1, the destruction of organic matter with sodium hypochlorite does not give oxalate and is consequently more suitable in this particular case – note that oxidation of organic matter is accompanied by oxidation of mineral compounds like those of Fe2+ – decarbonation and deferrification in an acid medium can destroy minerals like amorphous silicates with precipitation of silica and elimination of Fe et al. (Fröhlich, 1980). 5.2.3 Brief guide to interpretation of the spectra General Principles In IR, transitions between the different energy levels are subject to rules of selection as absorption is linked to variation in the dipole moment of

IR Spectrometry

147

the molecules. In the case of polyatomic molecules, all predictable frequencies cannot be observed because energy levels are degenerated for example by symmetry in the molecule. It is thus, very difficult to accurately predict the frequencies of fundamental vibrations in complex structures like those of clays, though recent computer programs have enabled progress to be made. Tetrahedrons of silica and octahedral aluminium or magnesium form the basic units of clay minerals: a tetrahedral structure can produce four modes of vibrations, and an octahedral structure six, but these are not all active and can be modified by isomorphic substitutions or by the nature of the structural cations. The interpretation of a clay spectrum can be carried out after comparison with the spectrum of a “pure” substance of comparable nature in order to eliminate uncertainties caused by chemical variations in the composition and order–disorder state. The need for known standard spectra of reference implies each laboratory should record all results of studies on soil minerals to be used as references in addition to consulting available data bases. Qualitatively, it is first necessary to locate the intensity of the diagnostic absorption bands of minerals and to assign them to molecular groups and possibly to types of precise vibrations. The degree of sensitivity is satisfactory in the case of certain minerals that have intense bands (e.g. kaolinite, quartz, gibbsite, calcite). Quantities of the order of 1% can be detected. For example, pure hydroxides and oxyhydroxides have a protonic environment that results in net vibrations of specific stretching; on the other hand, in the case of 2:1 and 2:1:1 minerals where chemical variations and isomorphic substitutions are frequent, displacement of the bands can occur; in this case it is useful to simultaneously use XRD analysis (cf. Chap. 4) and chemical analysis by selective dissolution (cf. Chap. 6) for secondary compounds of the soil, making it possible to draw up (silica of tetrahedrons) ratios that satisfactorily reflect the environment (alumina of octahedrons) of hydroxyls. 1:1 kaolinite has a SiO2-to-Al2O3 ratio of 2 and presents surface hydroxyls that produce four frequencies of characteristic IR absorptions (Table 5.1). Smectites and micas have a ratio of approximately 3. As hydroxyls in internal positions are associated with different octahedral cations, the absorption bands are not uniform in the 3,600 cm-1 area and displacement of the frequencies of hydroxyl stretching may be observed. The level of occupation of the octahedral sites (di- and tri-octahedral) can be determined by XRD analysis at line 060, but IR analysis can provide additional information.

148

Mineralogical Analysis

In the case of tri-octahedral minerals, the three sites are occupied and the axis of the OH bond of internal hydroxyl is perpendicular to the 001 reticular plane of clays, whereas in di-octahedral minerals only two sites are occupied; the proton of internal hydroxyl is pushed back towards the empty octahedral sites and the spectrum is consequently deformed. Procedure The unknown spectra are manually broken up into fields and the bands of maximum intensity are selected along with the wavenumbers of the maxima to compare with reference data. This search can be automated with computerized data bases but in practice also requires the use of laboratory reference data and continuous consultation of the literature. In solid minerals, interpretation is mainly based on the frequencies of the external molecular group, as the detection of vibrations of the internal crystal is only significant in far IR. IR spectrometry accounts satisfactorily for the molecular groups in which the atoms are in a specific environment. Molecular structures with characteristic bands can be isolated. IR Absorption Bands in Phyllosilicates The apparent simplicity of soil minerals masks the complexity of absorption bands of the fundamental modes of vibration (Fig. 5.6). The bands are displaced as a function of the crystalline environment, of substitutions, etc. The frequency and assignment of the bands require very precise spectra to separate slight variations from phases that are often about 2 cm–1. For example, distinction of amorphous silica, opal, biogenic silica by means of the Si–O, Si–O–Si, Si–OH vibrations are of this order of magnitude. – protons in hydroxyls groups, even if there is no long-distance molecular structure, which is useful in the case of “amorphous” substances – the active modes of tetrahedrons and octahedrons: OH oscillation, dichroism of OH sites in di-octahedral minerals

In the 3,700– 3,400 cm–1 and 950–600 cm–1 zones

IR Spectrometry

149

– displacement of the absorption bands as a function of the nature of the octahedral cations, effect of exchange (e.g. interaction of structural OH with interlayer cations)

– separation of the stretching vibrations of nonbound water allows better distinction of the halloysites disturbed by interfoliaceous water molecules
– vibrations of the silicate anion that are considered as the spectral print of clays, slightly coupled with vibrations of other structures (silica-oxygen bonds), Si–O stretching towards 1,000 cm–1, Si–O deformation towards 500 cm–1; – isomorphic substitutions of tetra and octahedral minerals can cause bands in far IR In the <400 cm–1 zone – Al3+ of octahedrons and oxygen bonds of adjacent layers – vibrations of exchangeable cations balancing the charges in interfoliaceous spaces of clays and substitution of exchangeable cations

In the 1,100–500 cm–1 zone

Taking the example of the adsorption bands of a common clay, 1:1 kaolinite (Table 5.1), it is possible to observe: – a similar configuration of the bands of proton vibration (stretching of external and internal hydroxyls) with respect to crystallinity; – a level of disorder that is detectable by the coalescence of the 3,6693,653 cm–1 doublet; – hydration water that is easily differentiated by deuteration and the appearance of a HOH band towards 1,630 cm–1

150

Mineralogical Analysis

– 1,110, 1,036 and 1,010 bands corresponding to stretching vibrations characteristic of SiO. In the case of Halloysites, widening of the bands can be observed between 3,700 and 3,600 (OH stretching) because of the structural distortion caused by variable hydration. Bands 795 and 758 cm–1 are of about the same intensity in kaolinite whereas in halloysites band 795 cm–1 is very weak.

Table 5.1. IR absorption bands of well-crystallized kaolinite Al2O3,2SiO2,2H2O, <2µm KBr pellet band –1 (s cm ) 3,695 3,669 3,653 3,619 3,410– 3,450
a

Vibrations

band –1 (s cm ) external OH external OH 690 538 470 430 364

Vibrations

deformation +

Mg/Al SiO +

stretching

external OH internal OH OH hydration

vibrations

Al/Mg octahedrons

1,630a

deformation

HOH hydration

345

miscellaneous

1,106– 1,112 1,036 1,010 938 912
a

stretching

out layer SiO

275

SiO... SiO... Deformation external OH internal OH

195

water of hydration (towards 3,450 cm–1) can be distinguished from the OH of structural hydroxyls by the presence of a HOH band towards 1,630 cm–1

IR Spectrometry
151

Fig. 5.6. Zones of assignment of the IR absorption bands in phyllosilicates (according to Stubican and Roy, 1961). Bands intensities: (t) 3+ 2+ very strong, (m) medium, (f) strong, (b) low ↓ changes in intensity → modification of frequencies by substitutions (X or X ) in tetrahedral sites (IV) and octahedral sites (VI)

152

Mineralogical Analysis

5.2.4 Quantitative Analysis Definition In mineralogy, improvements in quantitative analysis using IR have accompanied improvements in FTIR equipment, and quantitative analysis is an important tool for the study like pedogenic chronosequences, flows and sedimentary series (e.g. paleohydrology, paleoclimatology) or watershed functioning. It is always interesting to observe a mineral in its original state without pretreatment, with the exception of grinding in a non-aggressive liquid medium to reduce granulometry to 1–2 µm, then drying. The relations between the intensity of absorption of IR radiations and the concentration of mineral species are controlled by the law of Beer– Lambert which is applicable to solid media (Keller and Picket, 1949, or; Duyckaerts, 1959). However, this law is not always applicable to soil minerals for which the characteristic bands are not quite separate and often insufficiently homogeneous. The order–disorder states affect the band widths, and grain sizes influence band intensity. In this case, the limitations are obvious and the photometric response is seldom linear: the absorbance of a band is then no longer proportional to the concentration, and quantification cannot be correctly carried out as it can for gas or organic liquids. Since standard minerals (with structure, crystallinity, chemical composition and particle size similar to that of the samples) do not exist, only relative quantification can be obtained using this method, nevertheless it is possible to measure variations in the composition of minerals in a profile or toposequences with acceptable precision. For certain minerals with a relatively characteristic spectrum and a well defined base line such as some quartz, kaolinite, carbonates, the levels of detection will be about 1–2%. Purification by sedimentation allows concentration and simplification of the spectra. The state of crystalline order can cause significant variations, for example in kaolinites when structural OH between 3,700 and 3,600 cm–1 are used. The preparation of the sample is particularly important for quantitative analysis. Maximum absorption (without saturation) and minimum dispersion are required. The isogranulometric size of the particles must be less than 2 µm. The homogeneity of the discs must be perfect, the components should be dried at each stage of preparation or storage, and the thickness of the sample must be constant at less than 1 mm.

IR Spectrometry

153

Rigorous standardization of the procedures makes it possible to optimize measurements of soil mineral components without depending completely on the degree of crystallinity of powders as in XRD (cf. Chap. 4). Some minerals with short-range organization that are “amorphous” to X-rays, for example some aluminosilicates, crypto crystalline compounds (e.g. allophane, imogolite), and some forms of silica and of iron (e.g. ferrihydrites) can be quantified in this way. Procedure Overview The procedures for qualitative analysis (cf. Sect. 5.2.1 and 5.2.2) are applicable here, and particular care should be taken to: – desiccate the samples and binders at 105°C for one hour before weighing precisely (10–6 g); a temperature of 40°C can be used for heat-sensitive substances – grind the clay samples to less than 1–2 µm with an agate mortar or preferably with a tightly adjusted cooled vibratory grinder with agate balls in wet medium (in the presence of ethanol or of acetone); after drying, all the particles should pass through a 0.1 mm sieve; fractions that are quantitatively isolated by ultracentrifugation (cf. Chap. 3) can also be used – weigh a quantity of clay allowing transmittance ranging between approximately 20% and 70% for the whole band range of the spectrum (qualitative tests make it possible to fix the exact proportions between 2 and 5 mg of sample for 1 g of binding agent); mix with the appropriate binding agent (often KBr) and homogenize with a vibratory mixer with agate balls for 5 min – using 1 g of the above mixture, make three discs with a diameter of 13 mm each by weighing the same quantity of mixture (300–330 mg) to obtain a constant optical pathway of one mm or less; transfer to the pelletizer under vacuum (Fig. 5.5) and apply a pressure of 10 tons cm–2 for 2–3 min; the discs should be transparent, smooth, show no surface defects, be of constant thickness and present regular dispersion of the clay particles in the binding agent (this should be checked under the microscope); they should be handled using a forceps and stored in a desiccator with phosphoric anhydride; prepare the calibration discs in the same way – measure the absorbance of the discs made of binding agent (for the blank assay), sample disks and pure or complex calibration discs made

154

Mineralogical Analysis

according to the laboratory reference, which enable calibrations with variable proportions of minerals. Sediments Fröhlich (1980, 1989) recommended grinding with an agate ball to less than 2 µm in a cooled inert liquid medium. The fineness of grinding should be checked under the microscope. Optimal dilution is around 0.25%. – Prepare 1 g of mixture: 2.5 mg of sediment (precision 10–5 to 10–5 g) and 997.5 mg of KBr – homogenize carefully with a vibrating grinder with an agate ball; – press 300 mg of the mixture for 2 min in a 13-mm diameter mould under vacuum.

Fig. 5.7. Determination of T0 :good base line (a, b), approximation (c).

IR Spectrometry

155

The thickness of the KBr1 disc should be 0.83 mm and represent the constant optical pathway in all measurements. The disc should be smooth and transparent and should contain 0.75 mg of sample. Calculations Any absorbance due to the thinner (KBr) must be subtracted from total absorbance (KBr+sample). The transmission (and by conversion absorbance A) of the substance is measured on the spectrum starting from the base line. This line is often difficult to define for complex mixtures (Fig. 5.7) and requires approximations (effect of matrix, interference between bands, etc.). The relative error is often about 5%, and can be improved for certain minerals.2 Calibration using “pure” minerals or mineral mixtures with a composition close to that of the samples makes it possible to plot curves of absorbance = f (mass of mineral). One gram of mixture makes it possible to estimate repeatability on three discs weighing 300 mg. Remarks Strong orientation of minerals in the disc can generate errors. In 2:1 clays, variations in intensity of the stretching bands of hydroxyls can result from tri-octahedral components. With micas oriented perpendicular to the beam, only modes of vibration parallel to plane b will appear (Phlogopite). Conversely, in kaolinite, the intensity of the 3,619 cm–1 band is independent of the orientation (internal hydroxyl directed towards the vacant octahedral position). If titrations are carried out with a traditional dispersive apparatus, the resolution can be improved by finer slits but the energy will be weaker. As the width of the slits is not constant throughout the spectrum, care should be taken that the slits are not too wide, because the signals could be deformed and the law of Beer–Lambert would then not apply.

1 2

Density of KBr: 2.75 In spite of the use of reference minerals, the great variability of the absorption bands as a function of the chemical structure often results in insurmountable difficulties in calibration of 2:1 and 2:1:1 minerals. In this case, it is possible to obtain only semi-quantitative measurements that allow identification of changes in a profile, the mineral tracer being used as standard of comparison at a given level.

156

Mineralogical Analysis

5.3. Other IR techniques

5.3.1 Near-infrared spectrometry (NIRS) Principle Vibrations of light atoms that have strong molecular bonds with protons (N, C or O) are used to analyse organo-mineral compounds or organic matter. When the bonds are weak and the atoms are heavy, it can be difficult to detect and quantify the vibration phenomena. Wide bands are reproducible but are influenced by penetration of the radiation, and thus sensitive to the size of the particles and to moisture. Fine grinding is usually required to obtain particles of the same size and to reduce the background noise as much as possible. But acceptable results can be obtained with materials that are not finely ground (D. Brunet, IRD Montpellier, France, personal communication). Bond vibrations cause a response that depends on the number of molecules present and on their environment, this response then enables quantification The bands in the near-infrared field are more widely spaced than in medium and far IR, which limits the phenomena of overlapping. The first derivative of the signal can be used to improve precision. Measurements are made either by transmission–absorption, or by diffuse emission–reflectance (NIRA-DRIFT)3 on powders using wavelengths ranging from 1,000 to 2,500 nm (wavenumbers from 10,000 to 4,000 cm–1) in certain cases such as soil litter (water, protein content, total nitrogen, sugars, etc.), or on liquids using immersed optical fibres. Material Measuring equipment with diffuse reflectance IR is somewhat different from IR spectrometry using the transmission–absorption mode (cf. diagram in Fig. 5.8). The optical elements are made of quartz. They allow the complete near-IR spectrum to be acquired in a few seconds.

3

NIRS = near-infrared reflectance spectrometry NIRA = near-infrared analysis DRIFT = diffuse reflectance IR with fourier transform

IR Spectrometry

157

Method This non-destructive method requires fine grinding, but does not require a reagent, or weighing or measurement of volume. Measurements are rapid (≅30 s) and the unit cost of the measurement is low. Measurements depend on the physical factors that affect reflectance, i.e. particle size (reflectance, refraction, diffraction), their distribution (heterogeneity) and the distribution of the vacuums (compaction and induced orientations).

Fig. 5.8. Diagram of a near-infrared spectrometer (NIRS) 1: oscillating mirror allowing adjustment of the incidental beam on the sample and improving reflectance on the walls of the integration sphere. All possible angles must be represented starting from the normal. The total flow from the source that excites the sample is concentrated by quartz lenses. 2: Integration sphere allowing the effect of variations in particle size to be decreased with collection of the most intense radiation but rejection of specular components

Current NIRS systems have been greatly improved by progress in chemometric software which enables calculations that were previously impossible. Quantitative analysis is based on multivariate calibration using all the spectral information, not only absorbance at a given wavelengths but absorbance at all wavelengths of the spectrum. These calibrations use a wide range of methods of calculation, especially

158

Mineralogical Analysis

principal component analysis (PCA), regression on principal components (RPC), partial least square (PLS) regression and multiple linear regression (MLR). The software includes help in choosing the best method of calibration, even if the choice is still not always easy (Dardenne et al., 2000). This method is suitable for organic analysis but was extended to many different measurements on soils. Chang et al. (2001) used RPC calibration for FTIR determination of moisture content, total C, total N, CEC, sand content, silt content, clay content, macro-aggregation, potentially mineralisable N, C biomass, total respiration rate and basal respiration rate of soils. The only condition is the need to compile a database of soil references for calibration. For a given variable, calibration consists of (1) obtaining a measurement value by a reference method for all soil references (preferably including a wide range of concentrations of the given variable), (2) obtaining NIR spectra on the same soil references, (3) calculating and plotting the straight line of multivariate calibration with the value measured using the standard method in the x-coordinate and predicted value by NIRS in the y-coordinate. After calibration, the measurement of an unknown sample is very rapid: its real concentration in the x-coordinate can be deduced from its spectral data in the y-coordinate. But the apparently universal application of the method is not quite true. Calibration is always possible but not always significant (e.g. variability of the calibration curves obtained by Chang et al., 2001). Which soils to choose for the soil references (all soil types, or a given soil type)? What type of soil preparation? The complementary bibliography at the end of this chapter lists a few additional applications of NIRS in soil and litter studies. 5.3.2 Coupling Thermal Measurements and FTIR Spectrometry of Volatile Products Measurements are taken during TGA-DTA4 and enable determination of the nature of the gas products that appear during heat decomposition of the sample (EGD or EGA)5 The analyses are carried out by Fourier transform infra-red spectrometry (FTIR) by transmission or absorption in time of flight, as a function of the temperature and heating time. This dynamic technique enables real-time monitoring of the chemical or physicochemical conversions that take place during the rapid heating
4 5

DTA = Differential thermal Analyse, TGA = Thermo Gravimetric analyzes, cf. Chap. 7. EGD = Evolved Gas Detection, EGA = Evolved Gas Analysis.

IR Spectrometry

159

of the sample (controlled thermolyses or pyrolyses of organic or inorganic material is possible). A rise in temperature at moderate speed makes it possible to detect unstable radicals and molecular fragmentations by linking them to variations in mass and temperature (fusion, exo- and endothermic reactions, decomposition of mineralcarbonates, N, C, S, H compounds, oxidation, reduction, transfers of protons etc.). Pressure is a variable that affects sublimation and evaporation. Pressure can be modified by too rapid decomposition of an unstable product, but, depending on the temperature, can also cause molecular synthesis. Under low pressure, the most reactive gases diffuse quickly, avoiding possible recombination. Working under argon atmosphere at low pressure is generally recommended. But working under controlled atmosphere can highlight redox phenomena or, on the contrary, avoid them. Additional information can be collected by selecting a heating rate between 20 and 400°C min–1 depending on the speed of evacuation of the gas produced and on its detection or rapid titration before further gaseous reactions occur. 5.3.3 Infrared Microscopy FTIR analysis is possible on microsamples measuring from 20 to 500 µm. The IR microscope (cf. Chap. 8) consists of lenses with a Cassegrain mirror coupled with a high sensitivity MCT6 detector cooled with liquid nitrogen. It is possible to work with either transmission or reflectance. The resolution is approximately 8 cm–1 depending on the quality of the materials and the number of accumulations of spectra. It is also possible to use Raman spectrometry where the source of excitation is a monochromatic laser emitting in the red band to avoid the effects of fluorescence which occurs in the presence of certain organic materials. Quantities of the order of a pico and even of a femtogramme can be detected in this way. 5.3.4 Raman scattering spectroscopy Interest Raman spectroscopy has not been widely used in earth science because dispersive equipment is very expensive and the performance is often
6

MCT = mercury, cadmium, tellurium.

160

Mineralogical Analysis

insufficient due to the difficulty of obtaining a selection of wavelengths with high resolution. Progress in electronics has made it possible to design very sensitive detectors, very selective monochromators, and powerful monochromatic lasers, and to use FTIR spectrometers thus making this technique accessible and complementary to other IR spectrometries. Data processing enables very rapid treatment of the spectra. This technique makes it possible to supplement the information obtained in transmission–absorption IR spectrometry, as certain vibrations are only active in one of the two techniques, or their intensity differs because of the rules of selection. The symmetrical vibration bands are stronger in Raman spectroscopy and the asymmetrical vibrations are stronger in IR spectroscopy. The study of certain molecular structures that are difficult to differentiate such as rutiles, anatases and brookites is possible on microsamples. The nature of the chemical bonds and the orientation of OH groups can be determined without obstruction by any interstitial water that may be present. The method is not destructive and does not require complex preparation. It is possible to work on powder, even wet powder, which is impossible with other IR techniques. Principle Raman spectroscopy is based on the inelastic scattering of IR7 radiation with IR secondary emission at beat frequencies. The spectra are composed of fine lines which require a high resolution apparatus: – phenomena of fluorescence induced by the electronic transitions can disturb the spectra and mask the Raman signal if an excitation laser that emits in the visible spectrum (488 nm) is used; excitation by Nd:YAG laser emitting at 1,064 nm, the frequency corresponding to a zone of little occupied electronic transition, generally does not generate fluorescence, and coupling with a good-quality FTIR spectrometer eliminates difficulties due to insufficient resolution; – when a monochromatic radiation beam strikes a sample, a weak fraction of the re-emitted radiation displays modified frequencies that reflect the vibration frequencies of the sample. This fraction is measured in Raman spectrometry; the unchanged radiation fraction (Rayleigh elastic scattering) has to be removed by filtering.

7

Raman spectroscopy in visible or UV radiation can cause photodecomposition of the sample as well as thermal damage.

IR Spectrometry

161

Apparatus The basic apparatus is a FTIR spectrometer equipped with an interferometer coupled with data acquisition and processing software. It should have an external window to attach a Nd-YAG laser irradiation, a chamber for sample powders allowing irradiation modes of 90° and 180°, a spectral filtration module and if necessary a specific detector. A RamanFT microscope and accessories for Raman studies under very high pressure (diamond anvils) can be added to the FTIR spectrometer. The laser for excitation of the atoms and molecules must be monochromatic or adjustable in wavelength (for better selectivity). It must provide strong intensity (sensitivity of measurements), coherent radiation (spatial and temporal quality), and finally, if the beam is transported by optical fibre, low divergence. IR and Raman spectroscopy can be supplemented with other techniques for the study of structure (e.g. NMR, EXAFS). In spite of the development of these techniques, the IR and Raman methods remain competitive (ease of handling, reasonable cost of equipment) and allow a sufficiently detailed approach of the structure (including vacancies and substitutions, nature of the bonds in molecules) and of its consequences for rheology, for example, or for the study of pedogenesis phenomena that occur during weathering (e.g. coverings, interactions with the surface, adsorption of molecules). However, the quantification of the phases is often delicate if not impossible, because of the problems of orientation of clays and incomplete spectrum.

References
Chang C.-W., Laird DA, Mausbach M. et Hurburgh CRJr (2001) Near-Infrared Reflectance Spectroscopy-Principal Components regression analysis of soil properties. Soil Sci. Soc. Am. J., 65, 480–490 Dardenne P, Sinnaev G et Baeten V (2000) Multivariate calibration and chemometrics for near infrared spectroscopy: which method. Journal of Near Infrared Spectroscopy, 8, 229–237 Duyckaerts G (1959) The infra red analysis of solid substances. Analyst, 84, 201–214 Farmer VC et Palmieri F (1975) The characterization of soil minerals by Infrared spectroscopy. In: Soil components – 2 – Inorganic components, Gieseking JE ed., Springer, 573–670 Fröhlich F (1980) Néoformation de silicates ferrifères amorphes dans la sédimentation pélagique récente. Bull. Minéral., 103, 596–599 Fröhlich F (1989) Les silicates dans l’environnement pélagique de l’océan indien du cénozoïque. Mémoire Muséum National d’Histoire Naturelle, Paris, XLVI, 206p

162

Mineralogical Analysis

Keller WD et Pickett FE (1949) Absorption of IR radiation by powdered silice minerals. Am. Miner., 34, 855–868 Liu LG et Mernagh TP (1992) Phase transitions and Raman spectra of anatase and rutile at high pressures and room temperature. Eur. J. Mineral, 4, 45–22 Nail SL, White JL et Hem SL (1976) IR studies of development of order in aluminium hydroxide gels. J. Pharm. Sci., 65, 231–234 Nyquist RA et Kagel O (1971) Infrared spectra of inorganic compounds., Academic Press, New York Stubican V et Roy R (1961) Infrared spectra of layer silicates. J. Am. Ceram. Soc., 44, 625 Wada K (1966) Deuterium exchange of hydroxyl groups in allophane. Soil Sci. Plant Nutr., 12, 176–182 Weir CE Lippincott ER Van Valkenburg A et Bunting EN (1959) Infra-red studies in the 1 and 15 microns region to 30 000 atmospheres. J. Res. Natl. Bur. Stud., 63A, 55

Chronobibliography
Tuddenham WM et Lyon R.P (1960) Infrared techniques in the identification and measurement of minerals. Anal. Chem., 32, 1630–1634 Mitchell BD Farmer VC et Mc Hardy WJ (1964) Amorphous inorganic materials in soils. Academic Press. Adv. Agron., 16, 327–383 Hayashi H et Oinuma K (1965) Relationship between infrared absorption spectra in the region of 450–900 cm–1 and chemical composition of chlorite. Am. Miner., 50, 476–483 Hayashi H et Oinuma K (1967) Si–O absorption band near 1000 cm–1 OH absorption bands of chlorite. Am. Miner., 52, 1206–1210 Russell JD, McHardy WJ et Fraser A.R (1969) Imogolite: a unique aluminosilicate. Clay Miner., 8, 87–99 Wada K et Greenland DJ (1970) Selective dissolution and differential infrared spectroscopy for characterization of amorphous constituents in soil clays. Clay Miner., 8, 241–254 Conley RT (1972) Infra-red spectroscopy. Allyn-Bacon, 2nd. Edition Fieldes M, Furkert R.J et Wells N (1972) Rapid determination of constituants of whole soils using IR absorption. N. Z. J. Sci., 15, 615–627 Miller RGT et Stace BC (1972) Laboratory methods in Infrared spectroscopy., Heyden and Son Farmer VC (1974) The Infrared spectra of minerals. Minerals Sci. (London). Stepanov IS (1974) Interpretation of the IR spectra of soils. Pochvovedenie, 6, 76–88 Gadsden JA (1975) Infrared spectra of minerals and related inorganic compounds., Butterworth

IR Spectrometry

163

Griffiths PR (1975) Chemical infrared fourier transform spectroscopy., Wiley, New York Chemical Analysis, 43 Brame EG, Grasselli JG (1976) Infrared and Raman spectroscopy., Marcel Dekker, 1A White JL, Nail SL et Hem SL (1976) Infrared technique for distinguishing between amorphous and crystalline aluminium hydroxide phase. Proceedings. 7th Conference. clay Mineral Petrology (Czechoslovakia), 51–59 Marel HW, Van der et Beutelspacher H (1976) Atlas of infrared spectroscopy of clay minerals and their mixtures., Elsevier Amsterdam Proshina NV (1976) Use of infrared spectroscopy for identification of soil samples. Nauch. dokl. Vsshei Shk., Biol. Naudi, 3, 114–118 Brame EG, Grasselli JG (1977) Infrared and Raman spectroscopy., Marcel Dekker, 1B, 1C Hlavay J, Jonas K, Elek S et Inczedy J (1977) Characterization of the particle size and the cristallinity of certain minerals by infrared spectrophotometry and instrumental methods. I – Investigations on clay minerals. Clays Clay Miner., 25, 451–456 Hlavay J, Jonas K, Elek S et Inczedy J (1978) Characterization of the particle size and the crystallinity of certain minerals by infrared spectrophotometry and other instrumental methods. II-Investigation on quartz and feldspar. Clays Clay Miner., 26, 139–143 Ferraro JR et Basile LJ (1978) Fourier transform infrared spectroscopy. Applications to chemical systems., Academic, New York, vol. 1 Slonimskaya MV, Besson G, Dainyak LG, Tchoubar C et Drits VA (1978) Interpretation of the IR spectra of celadonites and glaucomites in the region of OH-streching frequencies. Clay Miner., 21, 377–388 Smith AL (1979) Applied infrared spectroscopy: fundamentals, techniques and analytical problem-solving., Wiley, New York, vol. 54 (chemical analysis). Farmer VC (1979) The role of infrared spectroscopy in a soil research institute: characterization of inorganic materials. Eur. Spectrosc. News, 25, 25–27 Ferraro JR et Basile LJ (1979) Fourier transform infrared spectroscopy. Applications to chemical systems., Academic, vol. 2 Hlavay J et Inczedy J (1979) Sources of error of quantitative determination of the solid crystalline minerals by inrared spectroscopy. Acta Chim., (Budapest), 102, 11–18 Olphen H Van et Fripiat JJ (1979) Data handbook for clay materials and other non-metallic minerals., Pergamon Martin AE (1980) Infrared interferometric spectrometers. In Vibrational spectra and structure, Durig J.R. ed., Elsevier, Amsterdam, vol. 8 Pouchert CJ (1981) The Aldrich library of infrared spectra., Aldrich Chemical Co, 1850 p Shika A, Osipova NN et Sokolova TA (1982) Feasibility of characterizing the mineralogical composition of soils by infrared spectrophotometry. Moscow Univer. Soil Sci. Bull., 37, 34–40

164

Mineralogical Analysis

Theng BKG, Russel M, Churchman GJ et Parfitt RL (1982) Surface properties of allophane, halloysite and imogolite. Clays Clay miner., 30, 143–149 Ferraro JR et Basile LJ (1983) Fourier transform infrared spectroscopy. Applications to chemical systems., Academic, New York , vol. 3 Fysh SA et Fredericks PM (1983) Fourier transform infrared studies of aluminous goethites and hematites. Clays clay Miner., 31, 377–382 Velde B (1983) Infra-red OH-stretch bands in potassic micas, talcs and saponites: influence of electronic configuration and site of charge compensation. Am. miner., 68, 1169–1173 Gillette PC et Koenig JL (1984) Objective criteria for absorbance subtraction. Appl. Spectrosc., 38, 334–337 Kosmas CS, Curi N, Bryant RB et Franzmeier DP (1984) Charactrization of iron oxide minerals by second-derivative visible spectroscopy. Soil Sci. Soc. Am. J., 48, 401–405 Prost R (1984) Etude par spectroscopie infra-rouge à basse température de groupes OH de structure de la kaolinite, de la dickite et de la nacrite. Agronomie, 4, 403–406 Kodama H (1985) Infrared spectra of minerals. Reference guide to identification and characterization of minerals for the study of soils. Res. Branch, Agric. Can. Tech. Bull., 1E Mulla DJ, Low PF et Roth CB (1985) Measurement of the specific surface area of clays by internal reflectance spectroscopy. Clays Clay Miner., 33, 391–396 Keller RJ (1986) The Sigma library of FT-IR spectra., Sigma chemical Co, vols. 1–2, 2894 p Griffiths et Haseth PR (1986) Fourier transform infrared spectrometry., Chemical Analysis Series, Vol. 83, Wiley New York, 672 p Russel JD (1987) Infrared spectroscopy of inorganic compounds. In Laboratory methods in infra-red spectroscopy, Willis H.ed., Wiley, New York Johannsen PG, Krobok MP et Holzapfel WB (1988) High-pressure FT-IR spectrometry., Bruker report, 39–43 Pouchert CJ (1989) The Aldrich library of FT-IR Spectra., Aldrich Chemical Co, vols. 1-3, 4800 p Mottana A et Burragato F (1990) Absorption spectroscopy in mineralogy., Elsevier, Amsterdam, Oxford, New York, Tokyo, 294 p Delvigne JE (1998) Atlas of Micromorphology of mineral alteration and weathering. The Canadian Mineralogist, special publication 3, Ottawa et IRD (ex-Orstom), Paris Silverstein RM et Webster FX (1998) Spectrometric Identification of organic compounds. Wiley New York, 482 p McHale JL (1999) Molecular spectroscopy., Prentice-Hall, London, Sydney, Toronto, 463 p Gillon D, Joffre R et Ibrahima A. (1999) Can litter decomposability be predicted by near infrared reflectance spectroscopy. Ecology, 80, 175– 186 Confalonieri M, Fornasier F, Ursino A, Boccardi F, Pintus B et Odoardi M (2001) The potential of near infrared reflectance spectroscopy as a tool

IR Spectrometry

165

for the chemical characterisation of agricultural soils. J. Near Infrared Spectrosc., 9, 123–131 Joffre R, gren GI, Gillon D et Bosatta E (2001) Organic matter quality in ecological studies: theory meets experiment. Oikos, 93, 451–458 Fearn T (2001) Standardisation and calibration transfer for near infrared instruments: a review. J. Near Infrared Spectrosc., 9, 229–244 Ludwig B et Khanna PK (2001) Use of near infrared spectroscopy to determine inorganic and organic carbon fractions in soil and litter. In Assessment methods for soil carbon, Lal R, Kimble JM, Follet RF et Stewart BA ed., Lewis, UK Ozaki Y, Sasic S et Jiang JH (2001) How can we unravel complicated near infrared spectra? – Recent progress in spectral analysis methods for resolution enhancement and band assignments in the near infrared region. J. Near Infrared Spectrosc., 9, 63–95 Reeves J B et McCarty G W (2001) Quantitative analysis of agricultural soils using near infrared reflectance spectroscopy and a fibre-optic probe. J. Near Infrared Spectrosc., 9, 1, 25–34 Tso, Ritchie GE, Gehrlein L et Ciurczak EW (2001) A general test method for the development, validation and routine use of disposable near infrared spectroscopic libraries. J. Near Infrared Spectrosc., 9, 165–184 Fidencio PH, Poppi RJ et de Andrade JC (2002) Determination of organic matter in soils using radial basis function networks and near infrared spectroscopy. Anal. Chem. Acta., 453, 125–134 Coûteaux MM, Berg B and Rovira P (2003) Near infrared reflectance spectroscopy for determination of organic matter fractions including microbial biomass in coniferous forest soils. Soil Biol. Biochem., 35, 1587–1600 Brown DJ, Shepherd KD, Walsh MG, Dewayne Mays M and Reinsch TG (2005). Global soil characterization with VNIR diffuse reflectance spectroscopy. Geoderma, doi:10.1016/j.geoderma.2005.04.025

6 Mineralogical Separation by Selective Dissolution

6.1 Introduction

6.1.1 Crystallinity of Clay Minerals Mineralogical characterization of cryptocrystalline minerals or minerals with short-range atomic arrangement (Fe, Al, Si, Mn, Ti, P) is essential to understand the geochemical and pedochemical phenomena that occur during the weathering of primary minerals, as well as to explain the evolution and the relative stability of the systems and the kinetics of chemical soil processes. These substances can represent the transition stage between the crystalline parent rock and secondary minerals, and are often regarded as tracers of evolution. The soil is an open system, i.e. it is able to exchange energy and matter with the outside. Most reactions occur under non-equilibrium conditions, and transitory states depend on aqueous or gas flows (chemical reactions at the solid–liquid and liquid– liquid interface), on relaxation times (particle diffusion, transfer of matter, etc.), and of course, on microbial activity. The individual accumulation of these substances, or their deposit in a fine layer of coating, modifies the activity of the structural sites of crystalline materials, and can inhibit the movement of ions, neutralize charges, or cause substitutions in the lattices. Gels, oxides and oxyhydroxydes, and aluminosilicates can develop charges (some of which are amphoteric). The high level of reactivity induced by their state of division allows adsorption of cations and anions. They can be neutralized by organic substances. These reactions confer greater resistance to weathering and to microbial action. Thus, the ultimate purpose of analysis of non-crystalline products may be soil genesis, and also:

168

Mineralogical Analysis

– Soil taxonomy (through processes of podzolisation, andosolisation, laterisation, etc.) – Soil mineralogy, mineralogical balances, purification before using other techniques (in particular methods that require the elimination of paramagnetic elements) e.g. ESR, EXAFS, Mossbauer, XRD, FTIR, SEM, EDX, WDX, TEM-HR, STEM1 – The study of the physical and chemical properties of the soil, studies of soil fertility (Fe deficiencies, P fixing, Al3+ toxicity, transport of heavy metals, destruction of the interparticle cements, aggregation factors, etc.) Identification of non-crystalline substances requires more than one method: XRD is not very useful with gels because if the quantity of gel is significant, or the early stages of development of a long-range crystalline structure are concerned, only broad bands will be obtained. Chemical dissolution methods are not sufficiently selective in mineralogy, as their action is based on acid, base, reducing, or complexing reagents, and consists for example in: – Breaking electrostatic (e.g. exchange reactions, Al3+ bridges) or coordination bonds (e.g. Fe3+ bridges) – Causing ionization of functional groups (organic matter) Dissolution must not only enable extraction of the different phases that are amorphous to X-ray but also: – Minimize chemical modifications (relative instability of the products to be extracted compared to the soil matrix, and avoid attacks of the clay lattices and primary minerals) – Limit hydrolysis of the extracted products – Avoid molecular rearrangements in the liquid phase (nucleation) – Maintain the extracted products in solution – Prevent the creation of chemical barriers (insoluble precipitate under the influence of the reagents) and the neo-formation of solid products

1

ESR electron spin resonance; EXAFS extended X-ray adsorption fine structure; XRD X-ray diffraction; FTIR Fourier transform infra-red; SEM scanning electron microscopy; EDX energy dispersive X-ray; WDX wavelength dispersive X-ray; STEM scanning transmission electron microscopy.

Selective Dissolution

169

All the extractions (e.g. single reagent or multiple reagents, single or sequential extractions) depend on thermodynamic constraints: – Ion activity, pH, concentration of the reagents, soil/reagent ratio, order of application of the reagents – Time factors, kinetics of extraction, stirring velocity, duration of contact, ageing of the gel – Temperature – Photolytic energy (UV catalyses of the chemical reactions) Initially the rate of mineralogical extraction is often significant, but subsequently levels off; it is also linked to the size of crystallites and perfection of crystallinity (defects, degree of disorder), or to the nature and the concentration of the elements in the liquid phase. Agreement with other studies, reproducibility and reliability will thus depend on the extraction procedures used. Unless justified by the need to adapt to specific problems, any modification in the procedure (proportion or concentration of reagents, time of contact, etc.) can cause serious errors in evaluation. The required degree of selectivity of mineral extraction can be obtained only by comparing different extractions that have been carefully purified by ultracentrifugation, and by chemical measurement (1) on the liquid phase containing the extracted products (congruent and incongruent reactions) and (2) on the solid phase (e.g. differential XRD, SEM, and EDX). Automatic calculation and interpretation can be performed with a limited number of reliable methods and this makes it possible to quantify each phase and to establish precise and reproducible geochemical balances. 6.1.2 Instrumental and Chemical Methods Measurement by X-ray diffraction works well for atomic lattices with long-range organization, but for substances with short-range arrangement without ordered superstructure, XRD gives flat, unusable spectra (Fig. 6.1). This is why substances presenting this type of flat X-ray spectrum are referred to as amorphous substances. Progress in instrumental methods has now made it possible to specify the nature of the phases and to determine their arrangement more precisely. The crystalline state is characterized by the periodic repetition of an atomic structure along three non-coplanar directions of space (Maziere 1978). The use of “non-crystalline” or crypto-crystalline substances, rather than paracristalline, is now allowed (amorphous, without structure; crypto, masked structure; para, almost a structure). It applies to the solids whose structure does not present a repetitive nature at long distance (molecular area at least 3 nm in diameter), but which

170

Mineralogical Analysis

presents a degree of order at short distance that confers specific properties. The short-range arrangement takes into account the mutual arrangement with the closest neighbouring atoms at the scale of interatomic distance. These substances do not have a clear spectral signature in X-ray diffraction. With another medium or at the longdistance scale, different types of non-crystallinity can also be observed: – Zones that present substitution disorders or structural dislocation: this is the case of amorphous substances in a well-arranged periodic structure (structure with defects); high-resolution phase contrast transmission microscopy and scanning transmission electron microscopy (STEM) in micro diffraction mode allow this type of arrangement to be detected and localised. – Extended zones without periodicity composed of clusters of randomly distributed particles with a short-distance arrangement; this is the case of gels of alumina, iron, silica, and some aluminosilicates (opaline, allophane-like, proto-allophane, allophane, proto-imogolite, gel-like, glass-like, vitric silica etc.). Some substances display the beginning of medium-distance organization (Fig. 6.1) which results in broad lines in XRD2 (e.g. imogolite, ferrihydrite, feroxihyte). Spectroscopic techniques can be used to study these minerals (see Fig. 6.2): – EXAF2 techniques provide accurate information on the interatomic distances of the closest neighbours and on the organization of the first layer of coordinance, but little information on relations between the polyhedrons (medium-distance coordination). – XANES2 techniques enable analysis of the sphere surrounding an atom, but a good knowledge of the structure is a precondition for success. – NMR2 spectrometry is selective for chemical structures but not very sensitive; the study of the hyperfine magnetic field of iron oxides makes it possible to measure the degree of crystallinity; in silica gels, the 29Si nucleus enables the different states of SiO4 bonds to be distinguished. Different forms of 31P can also be studied. 2 – ESR2 and ENDOR spectrometry enable analysis of the hyperfine and super-hyper-fine structures by electron spin resonance at the atomic scale.

2

See abbreviations p. 168. ENDOR Electron nuclear double resonance.

Selective Dissolution

171

Fig. 6.1. Crystalline and amorphous to X-ray compounds

172

Mineralogical Analysis

All these instrumental methods use radiations that correspond to ranges of distance suitable for the sub-micronic scales needed for the study of the structure of soil materials. The range of radiation extends from radio frequencies to X-rays and gamma radiations. Their high purchase price, the degree of specialization of the equipment and their use in very specialised analytical fields limits these types of studies to highly specialised laboratories.
g -

-

l n n

-

-

-

-

Fig. 6.2. Analyses of molecular structures using spectroscopic techniques (electromagnetic spectra and environmental probes; λ, wavelength; ν, frequency; n, wavenumber; radiations: F, far IR; M, medium IR; N, near IR, V, visible; UV, ultraviolet); NMR, nuclear magnetic resonance; ESR, electron spin resonance; XRF, X-ray fluorescence; XAFS, X-ray absorption fine-structure spectroscopy; XANES, X-ray absorption near edge structure; EXAFS, extended X-ray absorption fine structure

Instrumental methods allow observations at the sub-micronic scale, but chemical analysis enables analysis of entities that represent the average activity of certain particles because of their surface and their charge. Thus the continuum produced during weathering can be split in a satisfactory way by a series of extractions that make it possible to isolate compounds of increasing crystallinity that correspond, or not, to different chemical species. 6.1.3 Selective Dissolution Methods The range of “selective dissolution methods” is obviously limited because of the diversity of soil minerals (Table 6.1) and the difficulty involved in dissolving a well-defined single phase (Table 6.2).

Selective Dissolution

173

Dissolution depends on different factors: – The size of the “crystal” and the level of atomic disorder – Defects in stoechiometry or the blocking of active sites – The properties of the crystallographic faces (anisotropy) – The porosity of the systems, the density of surface defects, etc.

Fig. 6.3. Diagram of factors that control selective dissolution

The reagents used and possible pretreatments should not cause precipitation. They should ensure the maintenance in solution of the extracted products and if possible, limit recombinations in the liquid phase (Figs. 6.3 and 6.4). Oxides, hydroxides, and oxyhydroxydes cause dependent charges in the soil and their structures, bonds, surfaces, and reactivities vary with their degree of crystallinity and the degree of disorder of their lattices (Table 6.1).

174

Mineralogical Analysis

-

-

-

-

-

-

Fig. 6.4. Solubility of hydroxides as a function of pH and concentration (lower parts) and transformation of hydroxide gels (upper part)

6.1.4 Reagents and Synthetic Standards The complexity of iron forms and of non-crystalline products often requires the use of pure synthetic models of minerals with a crystallinity or a short-distance atomic arrangement that closely resembles the substances found in the soil. The precipitates should be prepared starting from products with a high degree of purity, as hydroxides tend to adsorb impurities because of their very great specific surface. Flocculation is achieved by adding H+ or OH– ions. Boiling causes transformation by dehydration and ensures the growth of the gel (nucleation). The time factor allows ageing of the gel, i.e. progressive slow crystallization (Fig. 6.4), transformation from a short-distance organization to an organization of a higher nature. For

Selective Dissolution

175

example, with aluminium in a ionic state, first precipitation of a monomer will be observed and then hydroxide: Al + 3 OH → Al(OH)3
3+ –

in different steps: Al(OH)2+→ Al(OH)2+ → Al2(OH)24+ etc. dehydratation is accompanied by loss of H+ during precipitation. With amphoteric aluminium compounds, the precipitates can be redissolved in alkaline medium forming soluble aluminates. All procedures must be strictly respected in order to obtain precipitation products that correspond to reproducible stages of formation. These procedures were defined by Henry (1958), Towe and Bradley (1967), Atkinson et al. (1968), Schwertmann and Taylor (1972; 1977), Murphy et al. (1976), Jeanroy (1983), Farmer and Fraser (1978), Pollard (1992), Lewis and Schwertmann (1979). Preparation of Iron Compounds Goethite – Dissolve 8.08 g of ferric nitrate (Fe(NO3)3,9H2O) in 80 mL water in a 250 mL Erlenmeyer flask. – Bring the pH to 7.5 by adding approximately 20 mL of 3 mol (KOH) L–1 solution drop by drop (while on a magnetic stirrer). – Leave for 3 h to form a deposit, siphon the supernatant and wash the precipitate four times with water to eliminate any soluble potassium nitrate that has formed. – Suspend in 100 mL water, then add 3 mol (KOH) L–1 solution to bring to a 0.3 mol (OH–) L–1 solution. – Store the solution in a polypropylene bottle at 20°C with occasional agitation for 2–5 years depending on the degree of nucleation desired. – Wash until elimination of KOH. – Dry in a ventilated drying oven at 50°C. Akaganeite – Grind a sample of pure FeCl2,4H2O in an agate mortar to pass through a 0.5 mm sieve. – Spread in a thin layer and allow hydrolysis to occur in contact with humid air for 6 months (the product will turn brown over time). – Wash with H2O to eliminate remaining Fe2+, then dry at 50°C.

176

Table 6.1. Main crystalline and non-crystalline oxides and oxyhydroxides in soil (Si, Fe, Al, Mn, Ti, mixed)

Fe

Al

Mn

Ti

Si

mixed gels (with charge) Si Al, Fe, Mn P

Fe Mn +2e Mn

3+

+e

Fe

2+

Al

3+

IV+

2+

Ti 4++ 2e Ti2+ Si+IV

hematite α-Fe2O3 hollandite αMnO2 cryptomelane αMnO2 pyrolusite βMnO2 birnessite δMnO2 ilmenite FeTiO3 manganite γMnOOH groutite αMnOOH brookite TiO2 α tridymite SiO2 α cristobalite SiO2 anatase TiO2 β quartz (high) SiO2 rutile TiO2

corundum αAl2O3(1)

α quartz (low) SiO2

allophane Al2O3–2SiO2,nH2O Imogolite

maghemite γFe2O3

gibbsite γAl(OH)3

magnetite Fe3O4 (Fe2+Fe3+O4)

bayerite αAl(OH)3(1)

hisingerite Fe2O3–2SiO2,nH2O penwithita SiO2–Mn evansite Al3PO4(OH)6,7H2O

goethite α-FeOOH

diaspore α-AlOOH

*lepidocrocite γFeOOH

boehmite γ-AlOOH

(leucoxene coesite see ilmenite) SiO2 stishovite SiO2

Mineralogical Analysis

ferrihydrite Fe2O3,2FeOOH

nordstrandite Al (OH)3

azovskite Fe3PO4(OH)6,7H2O

*feroxyhite δFeOOH *opal SiO2 * gel silica SiO2(nH2O) *gel titanium TiO2,nH2O biogene silica (SiO2) pH 1–3 chalcedony (fibrous) SiO2

*(kliachite Al2O3,nH2O)

hausmanite Mn3O4 (Mn2+ Mn3+ O4)

Selective Dissolution

akaganeite βFeOOH

todorokite

*stilpnosidérite (gel) Fe2O3,nH2O)

* vernadite δMnO2,nH2O

(limonite Fe2O3,nH2O see goethite)

lithiophorite (Al, Li) Mn O2 (OH)2

(3)

pH 7.5 Fe3+/2Fe2+

pH 4.0 + redisolve pH 9.0

pH 8.5–8.8

1 2 3 * ()

not very frequent or non pedogenic (primary minerals). isostructure. theoretical mean pH of precipitation in aqueous medium. gel, short-distance arrangement. obsolete terminology.

177

178

Mineralogical Analysis

Lepidocrocite – Dissolve 0.6 g of FeCl2,4H2O in 150 mL of a 0.2 mol (NaCl) L–1 solution saturated with nitrogen by bubbling using a peristaltic pump regulated at a flow rate of 15 mL min–1. – Agitate under nitrogen and adjust pH to 6.0 by adding NaOH 1 mol L–1 drop by drop. – Allow the pH to stabilize, then replace nitrogen bubbling by air; maintain the pH during oxidation (2 h 30 min). – Wash with water, then dry at 50°C. Poorly ordered Ferrihydrite – Dissolve 2.02 g of Fe(NO3)3,9H2O in 500 mL distilled water and bring the pH to 7.5 by adding 15 mL of 1 mol (NaOH) L–1 solution drop by drop. – Store for 18 h at pH 7.5. – Wash with water and dry at 50°C. Hematite – Prepare a solution M of ferric chloride (FeCl3,6H20). – Precipitate with NH4OH at 60°C. – Filter on rapid filter. – Wash until the Cl– test is negative (AgNO3 test). – Calcinate for 1 h at 500°C. Maghemite – Weigh 5 g of ferrous oxalate (FeC2O4,2H2O) and place it in a quartz crucible with a lid. – Heat gradually in an electric furnace at 410–420°C and maintain at this temperature for 1 h to eliminate water of constitution. The resulting product is close to maghemite. Organic amorphous iron – Extract the organic matter (OM) of a 20 g sample of podzolic soil with high humus content with 900 mL of 0.2 mol (NaOH) L–1 solution. – Centrifuge the extract, then filter (calculate the OM content expressed as C content). – Prepare a ferric nitrate solution (Fe(NO3)3,9H20) containing 108 g L–1. – In 225 mL of this solution (3.35 g of iron) add the desired proportion of OM extract. – Agitate while checking the pH with a pH meter; the final pH should be 5.0; a brown–red gel will precipitate. – Wash with distilled water until complete elimination of sodium. – Preserve the gel in water and check the iron content (mg mL–1).

Table 6.2. Estimation of extracted phases by differential selective dissolution (Tamm, ammonium oxalate reagent; DCB, Dithionite Citrate Bicarbonate reagent; Pyro, pyrophosphate reagent)

Selective Dissolution

Na2CO3 NaOH O.5 sequential mol L–1 9.8 >12.0 pH 1–14 +++ +++ + + 0 0 0 0 0 0 +++ +++ 0 +++ +++ +++ + +++ +++ +++ +++ + +++ +++ + ++ + +++ +++ + – – +++ – + +++ 0 +++ – +

alcaline NaF or KF Tiron

method, Tamm in DC tetraPyro extraction reagent darkness B borate pH of extracted phase 3.0 7.2 9.7 10.0 non-crystalline hydrated Al oxides +++ +++ + organic complexes +++ +++ +++ +++ crystalline compounds 0 + + 0 Si opaline-gels 0 0 0 crystalline compounds 0 0 0 ferrihydrite, feroxyhite Fe non-crystalline hydrated +++ +++ + oxides organic complexes +++ +++ +++ +++ crystalline compounds 0+ +++ 0 Mn non-crystalline oxides crystalline compounds allophane +++ + + imogolite ++ + + allophane-like comp. +++ +++ + (1)phyllosilicates 0+ 0+ 0 +++ strong to total dissolution, ++ medium or partial dissolution + weak dissolution 0 no significant dissolution + + ++ 0 +++ +++ +++ + +++ +++ +++ +(++) + + +++ +

+++ +++ +++ + strong pH increase with allophanes 179

180

Mineralogical Analysis

Preparation of Al3+ Compounds Boehmite – Prepare a 0.1 M solution of AlCl3. – Neutralize to pH 6 with 0.4 mol (NaHCO3) L–1 solution. – Leave in contact for one hour. – Bring to pH 8 by slowly adding 0.15 mol (NaOH) L–1 solution. – Store in a closed plastic (PTFE) bottle for 60 h at 160°C. – Cool, wash until elimination of bicarbonate and remove surplus sodium hydroxyde by centrifugation. – Store in suspension in closed polythene bottle. Preparation of mixed compounds Imogolite – Place 30 mmol of aluminium perchlorate in 2.5 L of deionised water. – Add 15 mmol of tetraethyl silicate, this corresponds to approximately 3.3 mL of commercial solution. – Homogenize, then bring the pH to 4.5 with soda. – The mixture will become opalescent; leave to stand overnight and the liquid will become clear. – Boil gently at reflux boiling point for 5 days. – Leave to cool, then add ammonia to gradually reach pH 9.0; the gel will precipitate. – Wash until elimination of sodium; store in water and measure the Al and Si concentrations.

6.2 Main Selective Dissolution Methods

6.2.1 Acid Oxalate Method Under Darkness (AOD) Principle This dissolution reagent is also called Tamm reagent. The method allows allophane and gels, iron, and aluminium organic complexes, hydrated oxides of iron, and aluminium (ferrihydrite, feroxyhite) to be dissolved. Imogolite is not completely dissolved in only one treatment. Phyllosilicates are only very slightly attacked, except if their level of

Selective Dissolution

181

disorder is significant. Lepidocrocite is sensitive to oxalate reagent. If several treatments (2–3) are performed; some Al and Fe crystalline compounds can be solubilized to a considerable extent. The ammonium oxalate-oxalic acid buffer induces processes of protonation, complexation, and reduction. In this way it can cause the transfer of protons, electrons and ions (Stum 1985; Furrer 1985–1987; Cornell and Schindler 1987; Schwertmann 1991). The oxalate ion forms three complexes with ferric iron:
2 Fe3+ + C2O 4− 2 Fe + 2C2O 4− Fe3+ + 3C2 O 2− 4 3+

⇔ ⇔

+ Fe C2 O 4

(6.1)
2

Fe (C2O 4 )

(6.2) (6.3)

3 Fe (C2 O 4 )3−

(ferrous iron also gives complexes with other constants of stability and solubility). An excess of oxalate buffer is required to bring the equilibrium reactions to stage (6.3); if not the C2O42– acceptors H+ and Fe3+ can induce competitive reactions, particularly if raising the pH decreases the constants of solubility. With Al 3+ a complexation of the following type occurs:
OH Al OH + -OOC -OOC Al

C C

O O

Below a pH of 3.5, the surfaces of the oxides are saturated with protons. The pH 3.0 zone is thus favourable because it controls charges below the point of zero charge (cf. Chap. 20). With respect to the reactional stages, protonation should be the first stage of dissolution of the compound, as this reaction allows better adsorption of the complexes and sumultaneous synergic action. In the case of non-crystalline compounds of iron and manganese, reduction is preponderant:

Fe3+ + e − ⇔

Fe 2+

182

Mineralogical Analysis

The complexing reagent has two effects: – Ferric ions become less oxidizing as the ferrous ions are more reducing; the Fe3+:Fe2+ ratio decreases, solubility increases; the reaction is autocatalytic – It affects pH (and maintains it at 3.0 with the buffer system) and thus influences the dissolution and stability of complexes; it prevents variations in the oxydoreduction potential which would otherwise be caused by the variations in pH. Changes in the Method Tamm (1922, 1931, 1934a,b) recommended an ammonium oxalate-oxalic acid buffer reagent to dissolve the inorganic gels including iron oxides, free silica, and alumina. This author suggested a pH of 3.25. A rapid examination of the composition of the reagents used since 1922 in the Tamm method highlights the many different procedures used (e.g. variations in the concentration of the reagents, in pH, in the soil/reagent ratio, contact time, agitation, temperature, photolysis, see Fig. 6.5). Some changes were made to adapt the method to the nature of the sample and its components. Comparisons with old data are often difficult because of complexity of the soil matrix and interactions, and especially because the exact operating conditions are unknown. Jung (1934) stated that the method was appropriate for light soils, but not for calcareous soils, and that it did not give repeatable results with heavy soils because of the attack of clays. Many studies have been published on different soil types using varying concentrations of reagents and a pH ranging from 3 to 6, or using other organic acids (tartric, citric, salicylic, benzoic, phtalic, malonic acid, etc.) combined or not with reduction with H2S, nascent hydrogen, or dithionite (Duchaufour and Souchier 1966). A significant stage in the development of the Tamm method was the discovery of photosensitivity during dissolution (Schoefield 1959) and of the effects of photosensitization (De Endredy 1963, this method being known as Tamm-UV), and especially the establishment of a reference procedure by Schwertmann (1964) also called Tamm reagent in darkness. This method is now used as an international standard. The time factor has a random influence on the reactions, the dissolution process generally being rapid at the beginning but tending to slow down considerably after 4 or 5 h. The influence of time is generally only critical if it is less than 2 h. Schwertmann fixed the average time at 2 h on the basis of a profile of traditional dissolution indicating the preferential dissolution of certain fractions (ferrihydrite, substances with

Selective Dissolution

183

short-distance arrangement, etc.). Mc Keague and Day (1966) showed that in their conditions, dissolution was better when the contact time was increased to 4 h (in darkness). Although a sequence of treatment of this duration can have a kinetic effect, only one extraction is the basis of this method.

Random results h

h

h

and h h h h h h h Higashi 1974

Fig. 6.5. Changes of the Tamm method using ammonium oxalate-oxalic acid reagent (basic procedures)

The soil/solution ratio generally has a limited influence on the results. Parfitt (1989) showed that extraction using the 0.15 M reagent at pH 3.0 with a soil/solution ratio of 1:100, agitation for 4 h at 20°C (in darkness) was satisfactory for many soils, but cannot be used if Al or extractable Fe exceeds 5%. In this case it is necessary to use a 0.20 M reagent and a ratio of 1:200. Extraction with 0.15 M oxalate or 0.20 M reagents at pH 3.0 gives equivalent results for allophane if the concentration of allophane does not introduce a limiting factor. The pH of extraction is critical. It controls the kinetics of dissolution of the crystalline and non-crystalline compounds. Maximum dissolution is reached at a pH of between 2.6 and 3.0, the protonation being synergistic with the reducing and chelating action of the oxalate reagent. If the pH

184

Mineralogical Analysis

rises above 4.0, the effectiveness of the buffer decreases drastically, the extracted quantities of iron decrease and selectivity is modified. The pH should thus be fixed at 3.0, to prevent possible variations. Temperature accelerates the reaction and modifies selectivity. The standard temperature is around 20°C. Preparation of the Reagents All the reagents should be prepared with reference products, bi-distilled water, or water deionised on a resin column that can fix Si. Acid ammonium oxalate: 0.2 M oxalate-pH 3.0 Tamm reagent – Dissolve 16.15 g of ammonium oxalate (COONH4)2,H2O3 and 10.90 g of oxalic acid (COOH)2,2H2O4 in approximately 900 mL of water; complete to 1000 mL. – Check the pH and bring it to pH 3.0 by adding ammonia or 0.2 M oxalic acid. – Prepare each week and store in a brown bottle protected from the light. – 0.2% superfloc (flocculation agent) in water (Cyanamid Corp.). – Matrix corrector for dilutions before atomic spectrometry: for 10,000 ppm K weigh 19 g of KCl, dissolve in approximately 900 mL of water, when the temperature of the solution reaches ambient temperature, complete to 1,000 mL. Procedure – Measure soil moisture on a separate sub-sample to determine the moisture correction factor (cf. Chap. 1). – On the laboratory balance, weigh 1 g of air-dried soil sieved with a 0.2 mm mesh (avoid over-grinding). – Put in a 100 mL bottle. – Add 50 mL of acid oxalate reagent; for soils with high extractableoxalate compounds (extracted Al or Fe compounds >2%) add 100 mL oxalate reagent and use a 250 mL bottle.

3

4

(NH4)2C2O4,H2O, mw: 142.12; can be awkward for certain clays (and can be replaced by sodium oxalate Na2C2O4); safety: a classified poison, do not ingest. COOH–COOH, 2H2O, mw: 126.07 decomposed by UV radiation; drying at 100°C involves losses by sublimation, decomposition at 160°C; safety: a classified poison – caustic – do not ingest).

Selective Dissolution

185

– Agitate 4 h in darkness. – Decant part of the supernatant in a 50 mL centrifugation tube. – Centrifuge for 10 min at 10,000 g; if the liquid is not perfectly limpid, resuspend, add 3 drops of superfloc and centrifuge again (two reactive blanks should be integrated into each series in addition to a soil standard of reference and two replicates on a sample of the series). After adequate dilution, analyses are carried out on this extract: – Si: ICP or AAS spectrometry at 251.6 nm, N2O/C2H2 flame – Fe: ICP or AAS spectrometry at 248.3 nm, air/C2H2 flame – Al: ICP or AAS at 309.3 nm, N2O/C2H2 flame – Mn: ICP or AAS at 279.5 nm, air/C2H2 flame and if necessary, Ti and P. If absorption spectrophotometry is required: – Destroy the oxalate matrix by boiling with concentrated nitric acid – Bring to dry and dissolve in 5 M hydrochloric acid – Evaporate to almost dry and dissolve in water, then complete to the required volume Calculations Data collected – P1, weight of wet or air-dried soil sample for measurement of moisture. – P2, weight of the soil sample dried at 105°C. – P, weight in mg of soil sample for extraction. – A, B,contents in the extract and the blank respectively (mg L–1). – D, dilution of the extract. – VR, mL of the oxalate reagent used for extraction. Moisture correction factor This measurement is essential to bring back the results to soils dried at 105°C, especially for all soils rich in non-crystalline substances like allophanic soils:

H = 100

P − P2 1 (%) P 1

Moisture correction factor = f = 100/H Calculation of contents of elements (Fe, Al, Si, Mn, Ti, P, etc.): % element = 0.1 (D VR f (A–B))/P Conversion factor of the content of an element to oxide content: % Fe2O3 = % Fe × 1.43 % MnO2 = % Mn × 1.58 % SiO2 = % Si × 2.14 % Al2O3 = % Al × 1.89

186

Mineralogical Analysis

Remarks The extraction is reliable enough for most soils. Identification of certain phases may sometimes be difficult when clays are disordered. Even when extracts are protected from the light, they can still undergo change. They should thus be analysed rapidly to avoid precipitations due to the instability of the reagent. The addition of superfloc is generally not necessary given the strong ionic force of the reagent and the resulting complexes. The supernatant liquid can be filtered by siphoning with a syringe equipped with a 0.45 nm filter (after decantation). Oxalate extraction is used in a number of different fields. – Pedology and pedogenesis, geochemistry (differential dissolutions and identification of the non-crystalline or little ordered phases, transitions between crystalline phases, chemical and methodological studies, effect on the soil structure, etc.) preferably using data from the dissolved phases. – Mineralogy uses the solubilized phases and the residual solid phases simultaneously for: – The study of substitutions, order–disorder states, compounds with short-range atomic arrangement; – For the preparation of samples (dissolution of cements between the particles, elimination of oxides to improve the intensity of diffraction of the crystalline compounds), to carry out differential XRD analyses (DXRD) and to enable analysis after elimination of paramagnetic compounds (Mössbauer, ESR, EXAFS spectrometries), to observe spatial distribution on thin sections of the soluble oxalate phases (Arocena 1988), and finally to model deterioration processes (resulting compounds, etc.). – Microbiology and agronomy to analyse biophysical and biochemical activity (effect on water retention, plasticity, availability for plants of active iron linked to oxalic acid contents generated in the soil (oxalic acid of biochemical origin causes the disruption of iron oxides). – In the case of calcareous soils, ammonium oxalate precipitates Ca++ cations in the form of calcium oxalate with solubility lower than 0.006 g per litre of water or acetic acid; the carbonate must thus be destroyed with the minimum quantity of acetic acid necessary before extracting ammonium oxalate and complementary measurements of the elements solubilized with acetic acid. – In reducing conditions (e.g. hydromorphic soils, histosols, andosols under permanent wet climate), the oxalate method cannot provide

Selective Dissolution

187

information on the initial state of oxidation of iron and manganese in the soil before extraction. – In andic soils and andosols–andisols the oxalate method can extract as much as the DCB method (cf. Sect. 6.2.2); the organic complexes of iron are dissolved; allophane, which is extracted after 2–4 h agitation with oxalate, can be estimated using the values for extracted silicon based on the hypothesis of the prevalence of Si–O–Al bonds; ferrihydrite can be estimated starting from extracted iron.

6.2.2 Dithionite-Citrate-Bicarbonate Method (DCB) Principle This method (Mehra-Jackson 1959–60) makes it possible to solubilize pedogenic oxides and hydroxides: – crystalline iron oxides (hematite, goethite), non-crystalline iron oxides and iron and aluminium organic complexes, as well as exchangeable iron and manganese oxides, some non-crystalline compounds with a SiO2:Al2O3 ratio of less than 0.5; – magnetite and ilmenite are only slightly attacked, as are gibbsite and allophane-imogolite aluminosilicates; however, magnetite can be significantly solubilized in certain cases (magnetite is strongly oxidized into maghemite in very oxygenated medium); – clays are not affected, but any iron present in the lattice of vermiculites and non-tronite can be significantly solubilized, particularly if the extraction pH is lowered; Reduction is the predominant process of this method, dithionite being a very active reducer (Deb 1950) below pH 9–10. Biologically reducible elements like iron or manganese are reduced and maintained in solution by complexation with the citric acid in the system buffered at pH 7.3 with sodium bicarbonate. The optimum pH for reduction is 7–8. Below pH 6.5, colloidal sulphur can precipitate resulting in a suspension in the extracts that prevents measurement by absorption spectrometry. The use of buffered medium limits this phenomenon. As the dithionite solution rapidly loses its reduction properties, complexation avoids reoxidation as well as the precipitation of iron sulphide and allows maintenance in solution of the extracted phases as long as the extraction time does not exceed 15 min. Initially, only one addition of dithionite was performed on the first extract. Subsequently, as a result of international influences and on the

188

Mineralogical Analysis

recommendation of the initiator of the method, two additions of dithionite were recommended. If required, two or three successive extractions can be performed to include crystallized iron compounds of relatively significant size. In this case, the extracts can either be mixed before analysis or analysed individually to measure a kinetic evolution of the solubility, summation is only carried out after each measurement. The temperature should be set at 75°C to accelerate the reaction and to limit the appearance of colloidal sulphur and iron sulphide, but also to minimize dithionite decomposition. This temperature should not be exceeded, and localised overheating should be avoided by using a waterbath. The iron contents should not exceed 0.5 g Fe2O3 in order to obtain an excess of reducer and complexant. The reaction of iron reduction in a slightly basic medium can be written: S2O42– + 4 OH– → 2 SO32– + 2 H2O + 2 e– 2 Fe3+ + 2 e– → 2 Fe2+ and in citric complexing solution S2O42– + Fe2O3 + 2 HOC(COO)33– + 2 H+ → 2 SO32– + 2 FeII– HOC(COO)3– + H2O The weight of the sample must be between 1 and 5 g without modifying the composition of the reagent (if iron cannot be complexed due to insufficient citrate, precipitation of black iron sulphide may occur). A buffered medium is used to avoid a change in pH resulting in variations in the Redox potential, each Fe3+ requiring two OH– during reduction. Certain authors tested reduction methods using dithionite in citrate with variable pH (Homgren 1967; Avery and Bascom 1982) or Tamm reagent (Duchaufour and Souchier 1966; Hétier and Jeanroy 1973; Loveland and Bullock 1976) or in other buffered and complexing mediums such as sodium tartrate-acetate (Deb 1950) or in a medium with a basic pH such as pyrophosphate (Franzmeier et al. 1965). Complexes with citric acid (a tridentate sequestering agent) are similar to those formed with oxalic acid and give very stable iron and aluminium compounds. Aluminium citric acid complexes have a stability constant, log K1 = 7.37. In the natural environment, the presence of citrate prevents, or delays, the precipitation of aluminium, as the sites of coordination are occupied by citrate, its hydrolysis is slowed down:

Selective Dissolution

189

Replacement of the water molecules and blocking of the sites of coordination occurs in the DCB extraction medium, which has a high concentration of citric acid. Hydrolysis thus becomes impossible (Kwong and Huang 1979). Some minerals, for example pseudo-boehmite, present a particular affinity (Cambier and Sposito 1991). Two other reagents (hydroxylamine hydrochloride and acidified hydrogen peroxide) were found more efficient than DCB method for selective dissolution of manganese oxides (Neaman et al. 2004a,b). The DCB treatment can cause structural disorders which can be observed by XRD or electron micro-diffraction. The adsorption of citric groups on certain clays can considerably slow down the departure of interfolayer water. This should be taken into account in the analysis of the residue containing the forms known as “free”. It should be noted that under these conditions, proto-imogolite cannot be transformed into the better structured imogolite. Preparation of the Reagents All the reagents should be reference products for analysis, water should be bidistilled or possibly deionized on resins suitable for the elimination of silica. – Sodium dithionite in powder form depending on the number of analyses, only small bottles of the product should be used in order to always have fresh product available. – Citrate-bicarbonate buffer: dissolve in distilled water before use 79.40 g of trisodium citrate (C6H5Na3O7,2H2O), 9.24 grams of sodium bicarbonate (NaHCO3), check the pH which should be 7.3 approximately, bring to 1 L. – Flocculation use either 400 g saturated sodium chloride, NaCl in 1 L of water, or 375 g saturated potassium chloride, KCl in 1 L of water, if further measurements by atomic absorption or ICP spectrometry are required, potassium chloride should be used in order to avoid stronger sodium concentrations, acetone.

190

Mineralogical Analysis

Procedure – In a flat-bottomed 100 mL centrifugation tube made of polypropylene or PTFE, place from 1 to 5 g of soil (0.2 mm particle size ) depending on the estimated concentration of ferric oxide (carbonates, organic matter and soluble salts must be removed from the sample beforehand). – Add a bar magnet and 45 mL of buffered citrate-bicarbonate reagent by means of a fraction distributor equipped with a PTFE syringe. – Place on a immersed magnetic stirrer in a water bath regulated at 75°C; when the sample reaches the temperature of the bath, using a measure, add 1 g of dithionite powder and continue to agitate at moderate speed for 5 min. – Add another gram of dithionite and agitate for 10 min. – After 15 min digestion, centrifuge for 5 min at 2,500 g to obtain a limpid solution (if the liquid is still cloudy, suspend and add a saturated solution of sodium or potassium chloride to cause flocculation then centrifuge again at 2,500 g; this treatment will make the centrifugation pellet more compact and thus complicate resuspension for a 2nd treatment; for soils originating from volcanic ash, it is often necessary to add 10 mL acetone before centrifugation to achieve satisfactory flocculation). – Decant the clear supernatant liquid in a 250 mL volumetric flask. – If the residue displays intense brown, black, or red colour, add 45 mL of buffered reagent and treat as above with two additions of dithionite and heat for 15 min at 75°C, re-suspend the compact centrifugation pellet to allow a homogeneous attack. – Centrifuge and decant in the same 250 mL flask (or analyze the second extract separately). – Wash the residue two or three times with 10 mL of buffered reagent, flocculate, centrifuge the rinsing products and add them to the previous extract. – Add 250 mL distilled water and homogenize. In each series, introduce two blanks (reagents only) and a reference sample. After adequate dilution (2–10 times) the filtrate containing free oxides and hydroxides should be analysed by atomic absorption spectrometry: – Al at 309.3 nm with an acetylene–nitrogen protoxide flame. – Fe at 248.3 nm with an acetylene–nitrogen protoxide flame. – Si at 251.6 nm with an acetylene–nitrogen protoxide flame. – Mn with 279.5 nm with an acetylene–air flame. – Ti, P, K, Mg can be also analysed if necessary.

Selective Dissolution

191

If colorimetry is used, certain methods make it possible to operate directly on the extracts, but it is preferable to destroy the buffered, chelating and reducing matrix by boiling with nitric or sulphuric acid and perhydrol. Iron is measured using 1–10 orthophenantrolin or ferron, aluminium using eryochrome cyanin, silica using molybdate taking phosphorous into account (cf. Chap. 31). Weigh the purified residue. The residue can be analysed using XRD, an instrumental method; the intensity of the lines is improved by DCB treatment (cf. Chap. 4); IR, ESR, NMR, EXAFS spectrometry (cf. Chap. 12); thermal analysis, DTA-TGA (cf. Chap. 7); or chemical analysis (total analysis, cf. Chap. 31); CEC (cf. Chap. 26); dissolution of aluminosilicates, etc.). Calculations Data collected – A, B: respective contents in the sample and blank extractions in mg L–1. – D: dilution factor. – f: moisture correction factor (cf. “Calculations” under “Acid oxalate Method under Darkness”). – P: weight of air-dried sample in mg. Calculations Oxide percentages should be calculated for all the elements: Fe2O3, Al2O3, SiO2, etc. The “weight of the initial sample” minus the “weight of residue” enables total free oxides and hydroxides to be calculated: Al, Fe, Si... % = 25 (A–B) D f/P See “Calculations” under Sect. 6.2.1 for the conversion factors of elements into oxides. Remarks This method gives reasonably reproducible results if the crystalline iron forms are sufficiently fine to offer enough surface area to allow a significant attack. Many different procedures have been proposed, but the current standard method is identical for the reagent concentration to that initially suggested by Mehra and Jackson (1959): 0.42 g of sodium bicarbonate and 3.52 g of sodium citrate, 2H2O in 45 mL of water. The main modifications one of the authors made of the method are: double reduction of dithionite on the same extract and the preparation of a single buffer-complexing reagent, which simplifies handling.

192

Mineralogical Analysis

The colour of the residue gives a good indication of the effectiveness of the treatment, but the presence of magnetite or ilmenite, which are not attacked by the DCB treatment, can colour the residue black or gray. The DCB method can be used to facilitate dispersion of clay whose suspension may be obstructed by pedogenic oxide and hydroxide coatings. The method of Holmgren (1967), whose reagent is composed of a rather unstable mixture of 17% sodium citrate and 1.7% sodium dithionite, is now sometimes used instead of the DCB method. It is considered to be equivalent to the Mehra–Jackson method, but simpler to implement and thus more suitable for repetitive analysis. Grinding to 0.2 mm allows a better attack of the iron forms present in concretions for example. Grinding is consequently generally performed to enable comparisons to be made, and in particular to compare the weight of the residues after treatment. Heating to temperatures above 80°C (or local overheatings) can cause the precipitation of black iron sulphide. In this case it is better to start the analysis again than to eliminate the precipitate with acetone and carbon tetrachloride. Dithionite treatment modifies the Fe3+:Fe2+ ratio. Ryan and Gschwend (1991) suggested replacing dithionite with titanium III in the citric-bicarbonate + ethylene diamine tetraacetic acid (EDTA) solution. This reduction method, with the very complexing reagent at a temperature of 80°C, should theoretically enable more complete dissolution of the amorphous ferric oxides and goethite. But hematite is less solubilized. Extraction is more easily achieved with the Ti (III) method than with the DCB method, which gives the Ti (III) method a more selective spectrum of dissolution. The behaviour of the aluminium compounds is different. The use of the cold Ti (III) method increases the degree of selectivity, but the titanium content of the soil must be low. 6.2.3 EDTA method Principle The EDTA (salt of Na) method according to Borggaard (1976), enables extraction of iron in an amorphous or very little ordered state as a result of biological deterioration (inorganic and organic non-crystalline iron) the ferrihydrite is dissolved. Both the water-extractable iron and the exchangeable iron must be in solution.

Selective Dissolution

193

This reagent is not suitable for the extraction of the amorphous and organic forms of aluminium because of the high pH of the extraction solution. The silicates and crystalline forms of iron and aluminium are not dissolved. The repeatability and selectivity of iron extraction are good. The method is reliable, but has one major disadvantage: the slowness of the extraction, the balance being achieved only after approximately three months of extraction. The most widely used process is hydrolysis and complexation in basic medium at ambient temperature (20°C). EDTA is an aminopolycarboxylic acid with six atoms suitable for the formation of chelates (from the Greek Khélé = crab grip) to which a metal cation is linked by coordination with the organic radical (e.g. two atoms of nitrogen and four carboxyl groups, see Fig. 6.6). The six positions around the metal (Fe2+) give a high complexing capacity and low selectivity, as most of the di- and trivalent elements are able to enter this type of complex. The constants of stability, log K at 20°C, are 13.8 for Mn2+, 14.3 for Fe2+, 16.1 for Al3+, 25.1 for Fe3+). In an alkaline solution (pH 10), 1:1 complexes are formed with pedogenic elements. An excess of complexing reagent is needed for satisfactory control of the dissolution rate . Depending on the pH, it may be possible to obtain H4Y, H3Y–, H2Y2–, HY3– forms (Y4– being the EDTA anion, see Fig. 6.6).


-

-

Fig. 6.6. Molecule of EDTA IV (top) and coordination complex with iron (bottom)

194

Mineralogical Analysis

In the extraction continuum, the extractable forms of iron with EDTA appear to be linked with the most active forms in pedogenesis, i.e. with the “free” least crystalline compounds with extensive surface contact and thus great reactivity. Both exchangeable iron and organic chelated forms are solubilized because the pH of the reagent is high and the EDTA complexes are stable. In agronomy, the availability of iron for plants (nutritional factor or possible chlorosis) is often checked using also two other complexing reagents (Lindsay-Norvell 1976): diethylene triaminopentaacetic acid (DTPA) or ethylene diamine di(o-hydroxyphenylacetic acid (EDDHA). The correlation between EDTA- and oxalate-extractable iron is good: the ferrihydrite is dissolved by EDTA or by oxalate, but soils containing hydroxyferric complexes can react differently with these reagents (Jeanroy 1983). The time factor is significant. The extraction profile is slow and dissolution is continuous up to around 90 days, when it stabilizes. The temperature is critical and attempts carried out to accelerate the reactions by raising the temperature to 75°C, as in the DCB method, resulted in unacceptable displacement of selectivity, some crystalline products becoming attackable and solubilizable. The effect of pH was tested by Borggaard (1976). It cannot be below 7.5. A pH of 10, similar to that used in the pyrophosphate and tetraborate methods, makes it possible to compare the organic phases extracted using the above methods and eliminates the effect of pH. But selectivity is random for aluminium, which, at a pH of 8–9, gives soluble aluminates. In EDTA, the concentration factor appears to have little influence on extraction, the kinetics of the reaction being controlled by hydrolysis. Intermittent agitation makes it possible to renew the reagent at the liquid– solid interfaces, thus avoiding the phenomena of local saturation. Clarification of the extracts is generally problem free. Preparation of the Reagents The initial procedure of Borggaard (1976) is a dynamic method. This author tested the effects of a concentration of the complexing reagent of between 0.01 and 0.1 M and a pH of between 7.5 and 10.5. This method is thus extremely long and cannot be adapted in its original form for repetitive analyses. A concentration of 0.1 M of EDTA at a pH of 10 is used as standard. This enables comparisons with the reagents which extract organomineral complexes at the same pH, i.e. pyrophosphate at pH 9.6–10 and tetraborate at pH 9.7. Commercially available EDTA is often in the form of sodium salt and is sold under different names: Versenate, Sequestrene,

Selective Dissolution

195

Titriplex II, Trilon B, etc. The empirical formula C10H16N2O8 corresponds to a molar mass of 292.25 g. Solution A: weigh approximately 29.225 g EDTA in 500 mL water. Soution B: dissolve 20 g NaOH in 250 mL water. Gradually mix B in A to bring the pH to 10. Complete to 1 L with distilled water. Procedure – On an analytical balance weigh precisely 2 g of soil ground to 0.2 mm in a 100 mL polypropylene or PTFE centrifuge tube. – Add 50 mL of 0.1 M EDTA reagent. – Stop the tube and agitate with the Vortex vibrator for 1 min. – Place the samples on mobile plates that can be used on an oscillating agitator and store the series protected from the light for 90 days with daily agitation for 5 min. – After 90 days of contact, centrifuge at 5,000 g for 5 min and filter. The extracted elements (mostly Fe, Al, Si, and P) are analysed by plasma emission or atomic absorption spectrometry. When absorption spectrocolorimetry is used, the EDTA matrix has to be destroyed (cf. “Procedure” under “Acid Oxalate Method Under Darkness”). In each series, introduce two blanks with only reagents and a reference sample. Calculations Data collected: A, B, D, f, P of the same type as in the preceding methods (cf. “Calculations” under section 6.2.1 and “Calculations” under “Dithionite - Citrate - Bicarbonate Method (DCB)” in this chapter). Contents Fe, Al, Mn, Si, P (%) = 5 (A–B) D f/P These contents are expressed as per cent of oxides (cf. “Calculations” under section 6.2.1). Remarks This method is one of the most reproducible and selective for amorphous iron and can also extract organometallic complexes if the pH exceeds nine. The extracted iron must be compared to the iron extracted by the oxalate reagent taking into account the fact that aluminous products dissolved at high pH can release iron for example resulting from isomorphic substitutions.

196

Mineralogical Analysis

It is useful to compare the results with the extraction of active iron available to plants (using DTPA or EDDHA reagents) to link the results to the phenomenon of chlorosis and to proceed from pedogenic observations to the identification of agronomic properties of the soil– plant–climate relationships. From an agronomic point of view, the extraction of phosphorous by EDTA is also useful to identify the proportions of P linked to Fe or Al which form part of the pool of “available P”. 6.2.4 Pyrophosphate Method Principle This method is used to analyze the forms of iron and aluminium complexed with the soil organic matter, in particular to differentiate the spodic and podzolic horizons where displacement of these complexes can be observed. Generally a good correlation can be obtained between extracted Al, Fe, and organic C. Well-crystallized iron oxides like goethite and hematite are not attacked and slightly ordered iron oxides are only slightly solubilized. The original procedure recommended by Alexandrova (1960), and given permanent form by McKeague (1967), cannot be applied as it stands but has to be modified with regard to the clarification of the extracts. The pyrophosphate anion P2O74– has chelating properties and can react with polyvalent cations to give insoluble compounds and soluble complexes with organic matter, for example: R(COO)4Ca2 + Na4P2O7 R(COONa)4 soluble + Ca2P2O7

But the complete mechanism of pyrophosphate action is not as clear as in the methods described earlier. The action of pyrophosphate has been questioned with regard to the organic forms of iron and aluminium: on one hand concerning the procedures and the reliability of the measurements, and on the other hand with regard to the mechanisms of extraction and the nature of the extracted products. Klamt (1985) mentioned “the denunciation of pyrophosphate extraction of Fe from soils (highly unreliable)”, indicating that an international consensus had not been not reached.

Selective Dissolution

197

With regard to the procedures The time factor of 16 h is regarded as critical, particularly when the results of this method are compared with those obtained by EDTA extraction after a contact time of 90 days. The pH, which was tested at different levels, is here fixed at 10.0 to enable comparison with the other extraction methods in basic solutions. Pyrophosphate has chelating properties. Originally the choice between the use of sodium or potassium pyrophosphate was more or less random. K pyrophosphate extracts slightly more than Na pyrophosphate and makes spectrometric measurements easier by avoiding strong Na+ concentrations, which are always awkward. But with certain clays, potassium has serious disadvantages5. Na pyrophosphate is consequently considered to give the best reproducible extraction and it is now used at a concentration of 0.1 M (Loveland and Digby 1984). At pH 10, pyrophosphate has peptizing properties which make the extracts very difficult to purify. Suspended particles are mainly responsible for the low rate of reproducibility and the lack of precision of the method, as the material in suspension is not chemically extractable by pyrophosphate. The efficiency of its centrifugation (speed and time of centrifugation) has been tested up to 100,000 g and compared to ultrafitration. A flocculating agent must be added before centrifugation, usually sodium sulphate at concentrations ranging between 0.25 and 1 M (Schuppli et al. 1983) or superfloc cyanamid N-100 (or Floerger Kemflock F 20 H). In this case the concentration is critical (Ballantyne et al. 1980). It is fixed at 0.2 mL superfloc for 50 mL of extract. Centrifugation at 20,000 g for 15 min results in clear extracts. Ultra filtration with 0.02 µm millipore filters can be used to eliminate any colloidal particles that may still be in suspension. In this case reproducibility is about 10–15%. With regard to the mechanisms of solubilization In the reaction between sodium pyrophosphate and soil, complexation cannot be the main mechanism, as iron linked to complex organic forms cannot be dissolved without solubilization of the amorphous forms of iron which are very reactive, as in the case of EDTA solubilization.
5 For example the K+ ion is specifically adsorbed by vermiculite or deteriored

micas because its diameter is compatible with the size of the adsorption sites. The selectivity of K+ with respect to Ca2+ or Mg2+ is increased by the hydroxy aluminous polymer deposits on the interfoliaceous surfaces. In the presence of a strong concentration of P, and in certain conditions, K and aluminous oxides together can result in the formation of taranakite.

198

Mineralogical Analysis

Compared to amorphous iron oxides, the iron pyrophosphate complex is not considered to be very stable at pH 10. Bruckert (1979) considered that sodium pyrophosphate shifts the organic matter of its complexes of coordination with the metallic sites of clays (ferric bridges) and is adsorbed instead of the humic compounds which are solubilized at an alkaline pH. Metallic compounds with a high charge, such as amorphous hydroxides, can behave in a similar way. All the complexes extracted by pyrophosphate comprise the “immovable complexes”. Micro-aggregates are destroyed and clayey and colloidal cements are dispersed. Fulvic acid-amorphous iron hydroxides complexes are extracted along with the organic molecules in the coatings on clays. Jeanroy (1981,1983) considered that dissolution induces a mechanism of peptization and solubilization. Adsorption of pyrophosphate on the soil particles increases the negative charges and increases their solublity in water. In contrast to EDTA, which has a flocculating effect, pyrophosphate puts ferruginous particles into suspension and these subsequently disperse in the extracts. Separation on millipore filter shows that in EDTA extracts iron is linked to small molecules and thus passes through the membrane. On the other hand, the pyrophosphate extracts contain compounds of greater molecular size that do not pass through the membrane, as only a small proportion of the chelated fraction is able to do so. Similarly ultracentrifugation of the EDTA extract does not separate the phases, clearly demonstrating that iron is in soluble chelate form, whereas pyrophosphate results in a significant colloidal centrifugation pellet. To summarize: with its peptizing action, pyrophosphate puts into suspension fine ferruginous particles, probably of ferric hydroxides, whose smoothness and small degree of atomic order are explained by the presence of organic matter which inhibits the crystallization of iron oxides. Bruckert’s “immovable complexes” appear to be mainly hydroxyferric complexes that reveal the preponderance of the mineral. From a practical point of view, pyrophosphate is effective only if the soil is under the influence of organic matter. In the opinion of Schuppli et al. (1983), the precipitation of iron in the pyrophosphate extracts could be due to the ageing of the extracts. Another dissolution mechanism could be the release by sodium pyrophosphate of small quantities of iron in the organic complexes, leaving these complexes negatively charged. They then become water soluble. The organic matter makes it possible to maintain quantities of iron in solution, but if the ratio reaches a certain level, precipitation will occur (Petersen 1976).

Selective Dissolution

199

If this description of the solubilization mechanism is accurate, pyrophosphate extraction could be a selective dissolution technique, especially if uncertainties concerning the purification of the extracts are overcome. From a practical point of view, pyrophosphate extraction enables the behaviour of certain soils to be differentiated for the purpose of classification, though without identifying the precise origin of the extracted iron; however organic forms are considered to be the most probable. Preparation of the Reagents Extraction 0.1 M sodium pyrophosphate: dissolve 44.6 g of Na4P2O7,10H2O in distilled water and bring to 1 L; check the pH which must be 10.0. Clarification Superfloc cyanamid N-100 (cyanamid Corp. Gosport, Hampshire, UK): dissolve 0.2 g of superfloc in 100 mL of water and agitate in darkness for 16 h with a PTFE magnetic bar stirrer, protect from the light in a brown bottle; a fresh solution should be made each week. Procedure With a laboratory precision balance, weigh one gram of soil (0.2 mm particle size) in a 250 mL polyethylene tube (with screw stopper). Add 100 mL of 0.1 M sodium pyrophosphate reagent and agitate for 16 h at ambient temperature (20°C). Add 0.2 mL of superfloc and homogenize on a rotary shaker for 10 min, then centrifuge at 20,000 g. With a millipore filter syringe remove from the quantities of supernatant needed to analyze the required elements (mostly Fe, Al, Si, and C). Add two blanks (with only reagents) and two reference samples in each series,. Measure the concentrations of Al, Fe, Si by plasma emission or atomic absorption spectrometry (cf. Chap. 31) with standards diluted in the extraction matrix. For measurements by absorption spectrocolorimetry, destroy the pyrophosphate matrix before measurement (cf. “Procedure” under “Acid Oxalate Method under Darkness”). Calculations Data collected: A, B, D, P, f as in other methods (cf. “Calculations” under Sect. 6.2.1).

200

Mineralogical Analysis

% contents (Fe, Al, Si) = 10 (A–B) D f /P The results can be expressed in oxides (cf. “Calculations” under Sect. 6.2.1). Remarks The method is well suited for differentiation of podzolic B horizons. The conditions governing centrifugation and clarification must be as homogeneous as possible, and the speed and time of centrifugation should be rigorously respected. If the series has to be stored before measurement by spectrography, store protected from the light in the refrigerator at 6–8°C. The percentages of extracted organic carbon are often measured on automated CHN apparatuses (cf. Chap. 10) at the same time as Fe, Al and Si are measured using spectrographic techniques (cf. Chap. 31). Series of extractions whose action is due to a gradual increase in the pH of the medium are often used. These differential extractions make it possible to characterize increasingly resistant forms: – A preliminary extraction using 0.1 M sodium tetraborate buffered at pH 9.76 enables the electrostatic linkages to be broken by simple exchange. The adsorbed organic molecules of low molecular weight are extracted. They contain complexed iron and aluminium that comprise the recently insolubilised “mobilizable complexes”. The soil aggregates are not destroyed (Bruckert 1970, 1974, 1979). – The 0.1 M pyrophosphate method at pH 10 then makes it possible to break the coordination linkages with the hydroxides and oxides in the coating on clays. – Lastly, the 0.1 M sodium hydroxide method at pH ≥ 12 (cf. Sect. 6.2.5) enables the organo-mineral linkages to be destroyed, even those of the allophane-humic acid complexes.

6

Preparation of reagent: 0.1 N sodium tetraborate pH = 9.7: dissolve 21 g of Na2B4O7,10H2O in approximately 900 mL of water; add 1.8 g of sodium hydroxide pellets; homogenize; check the pH which must be 9.7; bring to 1 L with deionised water. Extraction: 1 g of soil ground to 0.2 mm + 100 mL of 0.1 N sodium tetraborate solution; stir for 1 h; centrifuge at 20,000 g and continue as for pyrophosphate extracts (Al, Si, Fe, C, etc.).

Selective Dissolution

201

6.2.5 Extraction in strongly alkaline mediums Principle Methods using soda and sodium carbonate reagents are based on: – Dissolution in strongly alkaline medium of some silicon, aluminium, and aluminosilicate compounds; these can form soluble silicates and aluminates according to the simplified reaction: Al + NaOH + H2O → NaAlO2 + 3/2 H2. – concentration in the residue of insoluble compounds, and especially of iron and manganese. An attack using boiling 0.5 M sodium hydroxide solution for 2 min 30 s solubilizes organic forms of aluminium and silicon, hydrated noncrystalline and crystalline (gibbsite) aluminium oxides, opaline silica and diatoms, and finally amorphous or crypto-crystalline aluminosilicates like allophane and imogolite (SiO2:Al2O3 ratio = 1.5–2.3). Some 1:1 silicates are attacked and partially dissolved. Iron compounds are not extracted. An attack using 0.5 M sodium carbonate for 16 h at 20°C (Follet et al. 1965) has a more mitigated action and makes it possible to solubilize the organic and non-crystalline aluminium compounds as well as a certain proportion of gibbsite. Very finely divided siliceous compounds and opaline silica can be partially dissolved, amorphous aluminosilicates are solubilized, but allophane and imogolite are not completely solubilized. Phyllosilicates and iron compounds are not attacked. The iron compounds untouched by these two treatments can be studied in this enriched residue. But a more vigorous method with 5 M NaOH solutions and boiling for 2 h makes it possible to dissolve the majority of 1:1 clays and clay minerals present and thus ensure a higher concentration. Despite this treatment, some minerals such as quartz, anatase, rutile, cristobalite, and some 2:1 clays may still be present in the residue. The action of the reagents mainly depends on the dispersion of the particles, the state of division of the silicon and aluminium substances, the crystallinity and reactivity of the surfaces of ordered or X-ray amorphous compounds. Extraction in a 0.5 M NaOH medium with a limited period of boiling enables differential solubilization of aluminium and silicon compounds as well as of organic matter with high reactive surface, whereas wellcrystallized compounds are spared as their solubilization requires a much longer period of boiling. The solubility of aluminium hydroxides is amplified by an initial hydrolysis mechanism of the monomeric forms of aluminium (at low

202

Mineralogical Analysis

concentrations): AlOH2+, Al(OH)2+, Al(OH)3, Al(OH)4–, the latter existing only in an alkaline medium according to the reaction: Al3+ + 4 H2O → Al(OH)4– + 4 H+ Aluminium is in tetrahedral coordination. At higher concentrations, the polymeric forms gradually take the form of [Al6(OH)12,(H2O)12]6+ units. In addition, all the acid groups of organic macromolecules (humic and fulvic acids) are dissociated, and the polar and anion sites are easily solvated. Under these operating conditions, most of the organomineral bonds are broken, even the very resistant ones between allophanic-Al and humic acid of Andosols. The poorly ordered aluminium and silicon compounds or their organic and inorganic derivatives pass in solution in aluminate or silicate form. Dissolution in sodium hydroxide solution can result in an oxidative medium by breaking down some humic acid forms in the presence of oxygen, which modifies the Fe2+:Fe3+ ratios in the residues. Operating in nitrogen atmosphere can mitigate this phenomenon and also minimize the carbonation of sodium hydroxide by atmospheric CO2. Finally, in addition to the above processes, the strongly dispersing action of sodium hydroxide on the phyllosilicates also has to be considered. The elementary components of the stable micro-aggregates maintained in place by Fe, Al, Si oxide coating are first released by the dissolution process and then dispersed thereby increasing the action of the reagent at the solid–liquid interfaces. The alkaline character of the surfaces of the oxide decreases according to the series: Amorphous hydrated Al oxides 9.45 > γAlO(OH) boehmite 9.40 > αAl(OH)3 bayerite 9.20 > γAl(OH)3 gibbsite > γAl2O3 8.00

pH at isoelectric point

In addition to sodium hydroxide and carbonate, and sodium or potassium pyrophosphate, other reagents that have been used to extract the organic matter and organomineral complexes are: ethylene diamine (EDA), NN-dimethylformamide, sulfolane, pyridine, dimethyl sulfoxide, etc.

Selective Dissolution

203

Reagents The three mostly widely used reagents are described here as they correlate well with other selective extraction methods and characterize the resistance to the dissolution of the aluminous or siliceous products satisfactorily. Many alternatives have been proposed whose relative effectiveness (but not the degree of selectivity) is roughly classified in Table 3. Potassium hydroxide solutions are sometimes used instead of sodium hydroxide.
Table 6.3. Strongly alkaline reagents classified in order of decreasing efficiency 5 mol > 0.5 mol – – (NaOH) L 1 (NaOH) L 1 pH > 14 pH > 12 boiling boiling > 0.5 mol > 0.5 mol > 1.25 mol –1 – (Na2CO3)L 1 –1 (Na2CO3) L (NaOH) L pH 10.7 pH 10.7 80°C boiling 20°C

> 0.1 mol –1 (NaOH) L

20°C

2h

2.5 min

20 min (or shorter)

1 h (or variable times)

16 h

selective extraction of iron free oxides enrichment allophane imogolite

eliminates eliminates inter-particle gibbsite and cements, free oxides solubilizes 1:1 minerals

selective extraction of allophane imogolite

clay dispersion

Norrish and Hashimoto and Jackson Taylor (1960) (1961)

Follet et al. (1965)

Preparation – 5 mol (NaOH) L–1 solution: carefully dissolve 200 g of analytical grade sodium hydroxide pellets in 1 L of distilled water previously boiled to eliminate carbon dioxide Leave to cool with no contact with air and store in a polythene bottle. Prepare fresh solution every week.

204

Mineralogical Analysis

– 0.5 mol (NaOH) L–1 solution: dissolve 20 g of sodium hydroxide pellets in 1 L of previously boiled distilled water. Store in a polythene bottle. This reagent should be freshly prepared every day. – 0.5 mol (Na2CO3) L–1 solution: dissolve 53 g of anhydrous Na2CO3 in 1 L of distilled water and store in a polythene bottle. – 0.5 mol (HCl) L –1: take 42 mL of HCl d = 1.19 and bring to 1 L. Procedures Selective dissolution with 0.5 M Na2CO3 at 20°C (Follet et al. 1965): – Weigh 100 mg of soil sample ground to 0.2 mm and place in a 100 mL centrifuge tube. – Add 80 mL of 0.5 M Na2CO3 solution; close the tube and shake for 16 h with a rotary shaker. – Centrifuge at 5,000 g for 10 min. – Transfer the supernatant in a 200 mL volumetric flask. – Wash the centrifugation pellet with distilled water and recentrifuge. – Add to the previous extract and bring to 200 mL with deionised water. – Carry out spectrometric measurements on the extract without delay using atomic absorption or inductively coupled plasma emission (Si and Al). Selective dissolution with boiling 0.5 M NaOH (Hashimoto and Jackson 1960): – Weigh 100 mg of soil sample ground to 0.2 mm (or use the residue of a DCB extraction) in a 250 mL nickel or PTFE crucible. – Add 100 mL of 0.5 M NaOH boiling solution and homogenize. – Maintain boiling for 2 min 30 s. – Rapidly cool and transfer in a 250 mL polythene centrifuge tube and centrifuge for 10 min at 5,000 g – Transfer the supernatant in a 500 mL volumetric flask. – Wash the centrifugation pellet with distilled water, recentrifuge, and add the rinsing solution in the volumetric flask. – Adjust to volume with distilled water and measure Si and Al without delay using atomic absorption spectrometry or plasma emission spectrometry.

Selective Dissolution

205

Dissolution with boiling 5 M NaOH (2 h) (Norris and Taylor 1961): – Weigh 100 mg of sample ground to 0.2 mm (or use the residue of other selective extractions) in a stainless steel or PTFE beaker. – Add 100 mL of 5 mol (NaOH) L–1 solution. – Boil for 2 h. – Cool and centrifuge at 5,000 g for 10 min. – Decant the supernatant (which can be discarded or analysed). – Wash the centrifugation pellets with a little water, then wash three times with 0.5 mol (HCl) L–1 solution to dissolve the resulting sodalite and to eliminate sodium chloride, then wash again with water until negative reaction of chlorides. Dry the sample ready for the analysis of manganese and iron oxides. check the absence of kaolinite and sodalite by XRD. Calculations All the results are expressed in oxides (cf. “Calculations” under Sect. 6.2.1). Remarks The NaOH:Al ratio influences the development of aluminium hydroxides (Hsu 1977). It is thus important to standardize the procedures during the Al precipitation. Studies of 27Al by nuclear magnetic resonance show that, for a given concentration and pH, the nature of the gel gradually changes with ageing (Hsu 1984). These modifications, which occur in the natural environment, also occur in the synthetic mediums though at a different scale (Stol et al. 1976). NMR spectrometry is a particularly powerful tool to differentiate suspended or dissolved polynuclear species. Sometimes the modifications observed can explain such apparently induced variations by separation techniques such as centrifugation, ultrafiltration, or dialysis. Consequently it is better to carry out spectrometric measurements on products that have been recently extracted by different methods. For the 5 mol (NaOH) L–1 reagent, Kampf and Schwertmann (1982) recommend the boiling extraction method and in the presence of gibbsite to modify the reagent which must contain 0.2 mol (silica) L–1 in order to minimize phase changes and a possible increase in crystallinity, and to inhibit dissolution and recrystallization of substituted aluminium in goethites. Ferrihydrite, which can be converted into hematite and/or goethite, remains almost intact. However, the increase in the silica content results in more abundant precipitation of sodium and aluminium

206

Mineralogical Analysis

silicate (sodalite) which must be eliminated from the residue by several washings with 0.5 mol (HCl) L –1 solution. The method of concentration with 5 mol (NaOH) L–1 makes it possible to identify manganese oxides in the residue more clearly. These oxides are generally found at low concentrations and display weak crystallinity in soils. Birnessite and lithiophorite are found with iron oxides. Birnessite can be dissolved in hydroxylamine chloride as goethite and lithiophorite are not affected by this treatment. A final attack using the DCB method makes it possible to dissolve the goethite (Shuman 1982; Tokashiki et al. 1986). In the method based on boiling soda for 2 min 30 (cf. “Selective dissolution with boiling 0.5 M NaOH”), time is critical and must be carefully respected to avoid excessive solubilization of kaolinite and halloysite which can gradually pass in solution. The chlorites and montmorillonites, which were previously heated to 500°C, are not much affected. Spectrometric measurements should be made rapidly after extraction to avoid ageing of the extracts and precipitation of aluminosilicate. A large volume of solution compared to the weight of soil sample should be used to avoid the saturation of the extracts by aluminium and silicium. To avoid contamination, only PTFE, stainless steel, or nickel laboratory equipment is recommended. Pyrex glass can cause pollution by Si, Al, Fe. Dissolution of the gibbsite can be corrected if it is measured by DTA (cf. Chap. 7). Dissolution with 0.5 mol (NaOH) L–1 solution at boiling point does not enable very fine differentiation between amorphous and crystalline products because of the sensitivity of poorly ordered 1:1 clays and gibbsite. However, correlation with the oxalate method is generally good.

6.3 Other Methods, Improvements and Choices

6.3.1 Differential Sequential Methods Principle Many methods have been developed. When used in sequence, the methods described in Sect. 6.2 earlier can be considered as sequential multi-reagent methods.

Selective Dissolution

207

The Ségalen (1968) method uses an alternating process of hydrolysis and protonation in cold acid medium (without complexation or reduction) to solubilize some iron and aluminium compounds, then a treatment in hot alkaline medium (80°C) to solubilize the aluminium and silica compounds (cf. Sect. 6.2.5 earlier). These treatments are alternated several times to establish cumulative curves of quantity vs time (Fig. 6.7). Fractionation expresses the differences in the solubility of compounds in these mediums . Solubilization kinetics and hydrolysis constants depend on (1) the type, nature and concentration of the reagents, the soil/solution ratio and temperature and (2) nature and size of the sample, its elementary particle content, the degree of crystallinity and the extent of specific surface of clay minerals, the molecular arrangements and the degree of substitutions.
7

Fig. 6.7. Cumulative curve of extraction

The cumulative curve of the extracted compounds integrates these parameters (1) the lower part of the curve shows a parabolic segment that rises to a greater or lesser extent depending on the soil types which correspond mainly to the rapid dissolution of non-crystalline products
7

For example the dissolution kinetics of the iron forms in 0.5 mol (HCl) L–1 solution at 25°C (Shidhu et al. 1981) follows the series: (ferrihydrite, feroxyhite > lepidocrocite > magnetite > akaganeite > maghemite > hematite > goethite).

208

Mineralogical Analysis

with a very high specific surface area (Fig. 6.7); the kinetics of dissolution shows that a fraction of very fine crystalline products can also pass in solution in the same time interval; (2) the rectilinear upper segment has a weak slope indicating the weak dissolution of the wellcrystallized and chemically not very reactive products. A gradual change in the slope may indicate the end of dissolution of the very fine crystalline particles, or on the contrary, the appearance of practically non-ordered zones of preferential dissolution. Extrapolation of the upper segment at the intersection with the y-coordinate provides an estimate of the percent of amorphous substances. Reagents The reagents recommended by Ségalen (1968) for the acid attack are 2, 4, and 8 mol (HCl) L–1 hydrochloric acid solutions (the last concentration was finally retained) and for the alkaline treatment, boiling for 5 min with 0.5 mol (NaOH) L–1 solution. The extraction sequence comprises 8 alternating acid-alkaline treatments: – 8 mol (HCl) L–1: in a 1 L volumetric flask, add 830 mL of reagent grade hydrochloric acid to about 100 mL distilled water, agitate, leave to cool and bring to 1 L; alternatively if 6, 4, 2, or 0.5 mol (HCl) L–1 solutions are used, add respectively, 623, 415, 208, or 52 mL hydrochloric acid in the 1 L flask. – 0.5 mol (NaOH) L–1: add 20 g of NaOH pellets in a 1 L volumetric flask, dissolve in distilled water and bring to 1 L. The proposed modifications of this method concern the temperature (from ambient to boiling), the duration of contact and finally the use of only one reagent (2 mol (HCl) L–1) without alternating with NaOH. Procedure (Ségalen 1968): – Weigh 500 mg of soil (ground to 0.2 mm) in a 100 mL polyethylene centrifuge tube. – Add 50 mL of 8 mol (HCl) L–1 solution, homogenize and leave in contact for 30 min at room temperature. – Centrifuge at 5,000 g for 10 min. – Transfer the supernatant in a 100 mL volumetric flask. – Add 45 mL of deionised water, resuspend the centrifugation pellet by Vortex agitation. – Centrifuge at 5,000 g and add the washing water in the volumetric flask.

Selective Dissolution

209

– Complete to 100 mL; this is extract A (acid) for the measurement of iron, aluminium, and silicon. – Add 50 mL of NaOH 0.5 mol L–1 solution in the centrifuge tube and resuspend. – Place in the boiling water bath for 5 min. – Centrifuge at 5,000 g for 10 min. – Collect the supernatant in a 100 mL volumetric flask. – Wash the residue with 45 mL distilled water and add the washing water to the extract; complete to 100 mL; this is extract B (alkaline) for the measurement of alumina and silica. Repeat the double extraction process eight times. Calculations All the results obtained by atomic absorption or inductively-coupled plama emission spectrometry are expressed in per cent of oxides (cf. “Calculation” in Sect. 6.2.1). For alumina and silica, the results obtained on extracts A and B must be added. For iron, the acid solution is used alone. Cumulative curves of iron, alumina, and silica are established (Fig. 6.7); the tangent to the right of the curve gives the percentage of amorphous compounds. Remarks The procedure using the above concentrations (alternating 8 mol (HCl) L–1 and 0.5 mol (NaOH) L–1 with 5 min boiling), was found to be too energetic and insufficiently selective for many pedological situations and many minerals as the amorphous substances were often over-estimated. This method thus needs to be adapted to the type of soil being analysed. The original unmodified method dissolves amorphous iron and complexed iron, but if the use of very aggressive reagents results in the rapid dissolution of the amorphous substances, it also solubilizes crystallized products: 1:1 clays, mica, chlorite, biotite, hornblende, nontronite, gibbsite, etc. (Colmet-Daage et al. 1973; Quantin and Lamouroux 1974; Quantin et al. 1975; Yong et al. 1979; Bentley et al. 1978, 1980; Quigley et al. 1980, 1985; Torrance et al. 1986). The degree of selectivity may be insufficient for certain soils and amorphous substances will be considerably over-estimated. On their own, the kinetics curves do not show which forms are truly solubilized. The kinetics of dissolution is significantly correlated with the concentration of the reagents, and to time and temperature. Modifying these parameters can significantly change the dissolution kinetics and selectivity. The results always have to be linked with a precise procedure.

210

Mineralogical Analysis

As is true for all the other methods, an excess of reagent is necessary to avoid the effects of saturation. This method is not really suitable for repetitive analysis without robotization because of the length of the successive extractions. Six double extractions can be carried out each day on a small series. The study of goethite dissolution (Cornell et al. 1974, 1975, 1976; Schwertmann et al. 1985) in 0.5 mol (HCl) L–1 at 20 and 60°C using transmission electron microscopy (TEM) shows that preferential zones of dissolution create new surfaces that can give sigmoid curves of dissolution vs time. This mechanism of dissolution is related to the structure of the molecular configurations. The kinetics of dissolution of goethite is slowed down by a strong Al3+ substitution. These studies highlight the changes in the concepts of dissolution and of “amorphous” substances as well as the need to check the size and the state of surfaces of minerals using SEM and XPS microscopy. 6.3.2 Selective Methods for Amorphous Products Many other methods are used for the description of amorphous substances, in particular the release of hydroxyls by fluorides and trimethylsilylation. Release of Hydroxyls by Fluorides This method is based on the general reaction: (mineral-OH) + F → (mineral-F) + OH
– –

which satisfactorily accounts for the reactivity of the crypto-crystalline components, aluminosilicates like allophane or imogolite, or amorphous aluminium forms, the reactions being very rapid. Al (OH)3 + 6NaF → Na3AlF6 + 3NaOH The above reaction made it possible to develop a NaF field test for – andosols. The total released OH correlates well with the amorphous forms of Al and Si, with a weak influence of Fe. Quantification is performed by titration at constant pH: – Put 25 mg of soil ground to 0.2 mm in contact with 0.85 mol (NaF) L–1 solution solution at pH 6.8 and at 25°C;

Selective Dissolution

211

– Using a titrimeter, maintain the pH at 6.8 for 30 min by adding a titrated acid solution. The quantity of acid required for neutralization provides an estimate of the non-crystalline substances. The reaction can be performed stoechiometrically only in the absence of organic matter and carbonates. The crystalline substances are not sufficiently reactive to react significantly in such a short period of time. Trimethylsilylation This method is specifically used for silica compounds; it is derived from organic chemistry and is based on the reaction of the silicic acid with organo-silicic compounds which produce stable and volatile organosilyl derivatives that can be separated and identified by gas chromatography. The basic reaction can be written briefly:

This reaction makes it possible to determine the degree of polymerization of silica and to link it to different types of deterioration. Inorganic gels that are rich in aluminium are particularly reactive. Crystalline substances are not very active. Place 50 mg of soil sample in contact with 9 mL of hexamethyl disiloxane (HMDS) at 4°C in 0.2 mL of water and an internal standard (n-Eicosane, CH3(CH2)18CH3). After homogenisation, add 2 mL trimethylchlorosilane (TMCS)8 at 4°C. The hydrolysis releases hydrochloric acid which reacts with the soil amorphous silicates by exchange of cations and protonation. Leave in contact for 16 h at 4°C. The silicic acid forms trimethylsilyl derivatives.
8

8

TMCS, Trimethylchlorosilane (CH3)3SiCl; HMDS, Hexamethyldisiloxane (CH3)3Si–O–Si(CH3)3.

212

Mineralogical Analysis

Table 6.4. Approximate balance of soil mineral phases from different extractions Fe forms symbol description – analysis essential for the establishment of geochemical and mineralogical balances and characteristic ratios – the method of attack by boiling HCl is insufficient for satisfactory quantification. An alkaline fusion should be carried out. Minerals such as pyrites, marcasite, and magnetite are difficult to mineralize. ratios Fe(T)/Clays Fe(CBD)/Fe(T) are indices of weathering of the horizons. Degree of evolution of hydromorphic soils. iron extractable by DCB reagent (cf. Sect. 6.2.2): iron mobilizable in pedogenetic processes. This iron is of pedological interest since it is not linked with the structure of the silicate lattices (clays, primary minerals) that presents different hydroxylated and hydrated forms with a varying degree of crystallinity. ratios Fe(f)/Clays, Fe(oxda)/Fe(DCB) indicate the degree of weathering and crystallinity of free oxides and concretions. Fe(DCB)/Fe(T) is the index of deterioration and mobility of iron in a profile. Fe(DCB)–Fe(oxda) Fe(T)–Fe(DCB) the iron in silicates originating from primary minerals and the lattices of phyllitic minerals; a reduction indicates progression in deterioration (cambic horizon) total free Al (exchange acidity, exchangeable Al3+, ECEC, etc.), marker of deterioration (acidolyse, complexolyse) free Mn Al, Si, Mn, and others

total iron

Fe(T)

free iron

Fe(f) = Fe(DCB)

crystaline free iron

iron in silicates

Fe(sil)

Selective Dissolution
not wellcrystallized iron oxides wellcrystallized iron oxides Fe(OA) Fe(oxda)–Fe(EDTA)

213

Fe(OC)

Fe(DCB)–Fe(oxda) the well-crystallized iron oxides are identifiable using instrumental techniques (XRD, DTA, etc.). These oxides are inherited or newly formed. The presence of lepidocrocite and magnetite which are sensitive to the oxalate treatment may affect the results. Magnetite is inherited and is not formed during pedogenesis. immovable complex Al(tetra)/Al(Pyro) decreases with ageing of the gel (soluble aluminates in alkaline medium)

iron in amorphous forms and organic complexes complexed iron

Fe(pyro)

Fe(tetra)

mobile or mobilizable complex. Method used in sequence with reagents of increasing efficiency: Tetraborate pH 9.7<pyrophosphate pH 9.8 < NaOH pH 12 – Fe(tetra):Fe(pyro) characterizes the degree of evolution of complex Feorganic matter, if Fe(tetra) decreases and Fe(pyro) increases there is pedogenesis with strong biological deterioration. iron extracted by oxalate in darkness (cf. Sect. 6.2.1 above), if Fe(oxob), Fe(EDTA), Fe(pyro) are weak there is weak deterioration

Fe(oxda)

Add 2 mL H2O to hydrolize the excess of TMCS. Centrifuge to separate solid and liquid phases and transfer the organic solution on 1 g cation exchange resin (Amberlite 15). After 5 h of contact and filtration,

214

Mineralogical Analysis

separate the extract and quantify the trimethylsilyl derivatives by gas chromatography with nitrogen as carrier gas. 6.3.3 Brief overview of the use of the differential methods There are many publications concerning each method of extraction making it possible to find examples of the use of all the methods, either alone or in comparison with other methods, e.g. Aomine and Jackson (1959), Bruckert (1979), Jeanroy (1983), Baize (1988), KrasnodebskaOstrega et al. (2001). The final objective may simply be the elimination of substances (e.g. cementing agents) that obstruct dispersion or instrumental observation (e.g. XRD or IR spectra). In this case, it is not essential to identify the soluble products; on the contrary, to understand the mechanisms of mineralogical and pedogenic evolution both the soluble phases and the solid phases have to be taken into account. Comparisons should be made of results expressed in per cent of oxides; this is a practical way to homogeneously quantify the mineralogical phases but does not correspond to the chemical forms actually present in the soil. The combination of several methods makes it possible to establish an approximate balance of the different aluminium and iron forms in the soil (Table 6.4). Isolated entities are often not sufficiently specific and results are somewhat empirical, but give a precise enough view of the weathering phenomena. In general, variations in extracted contents of a given element (initially iron, then aluminium, silica, and manganese) are compared on the same sample first by total analysis and then by selective analyses (free forms, amorphous forms, etc.). Next, variations between horizons of a soil profile and finally variations between profiles can be compared. Correlations with size and behaviour of components can be highlighted. Iron is the main element used as a tracer of deterioration. It is insoluble in the ferric oxidation state and a loss generally indicates a pedogenic process of reduction, also indicated by the colour of the soil. The processes of biochemical humification result in the formation of iron–organic complexes. In andosols, Al and Si, components of allophane and imogolite (Fig. 6.1), are measured first. In these soils, allophanic materials can sometimes be estimated indirectly, for example by the measurement of the cation exchange capacity on a sample from which the organic matter has previously been eliminated; one part of the sample is treated with a

Selective Dissolution

215

sodium carbonate solution for 60 min, and another part with a boiling sodium acetate solution for 15 min, and the result will be a difference in cation exchange capacity that correlates well with allophane content (Aomine and Jackson 1959). An instrumental technique, differential diffractometry, makes it possible to evaluate the action of the dissolution reagents and the behaviour of soil materials. Several diffraction spectra of a sample are made before and after treatment by selective dissolution. An appropriate computer program can be used to calculate the difference between the spectrum of the treated sample and the spectrum of the untreated sample. Quartz or αalumina is used as an internal standard to calculate the scale factor for the subtraction of the spectra. This enables the spectrum of the dissolved substances to be reconstructed during selective treatment (Campbell and Schwertmann 1985).

References
Alexsandrova LN (1960) The use or pyrophosphate for isolating free humic substances and their organic-mineral compounds from the soil. Soviet Soil Sci., 190–197 Aomine S and Jackson ML (1959) Allophane determination in ando soils by cation exchange delta value. Soil Sci. Soc. Am. Proc., 23, 210–214 Atkinson RR, Posner AM and Quirk J (1968) Crystal nucleation in Fe (III) solutions and hydroxyde gels. J. Inorg. Nucl. Chem., 30, 2371–2381 Avery BW and Bascomb CL (1982) Soil Survey Laboratory Methods., Soil Survey of England-Wales (Harpenden), 6 Baize D (1988) Guide Des Analyses Courantes En Pédologie: Choix Expression - Présentation - Interprétation., INRA, 172 pages Ballantyne AKD, Anderson DW and Stonehouse HB (1980) Problems associated with extracting Fe and Al from saskatchewan soils by pyrophosphate and low speed centrifugation. Can. J. Soil Sci., 60, 141– 143 Bentley SP, Clark NJ and Smalley IJ (1980) Mineralogy of a Norwegian postglacial clay and some geotechnical implications. Can. Miner., 18, 535–547 Borggaard OK (1976) Selective extraction of amorphous iron oxide by EDTA from a mixture of amorphous iron oxide, goethite and hematite. J. Soil Sci., 27, 478–486

216

Mineralogical Analysis

Bruckert S (1979) Analyse des complexes organo-minéraux des sols. In Pédologie 2, constituants et propriétés du sol, Bonneau M. and Souchier B. ed. Mason, IX, 187–209 Cambier P and Sposito G (1991) Adsorption of citric acid by synthetic pseudoboehmite. Clays Clay Miner., 39, 369–374 Campbell AS and Schwertmann U (1985) Evaluation of selective dissolution extractants in soil chemistry and mineralogy by diferential X-Ray diffraction. Clay Miner., 20, 515–519 Colmet-Daage F, Gautheyrou J, Gautheyrou M and De Kimpe C (1973) Etude des sols à allophane dérivés de matériaux volcaniques des Antilles et d’Amérique latine à l’aide de techniques de dissolution différentielle. Ière partie. Etude des produits solubilisés. Cah. ORSTOM série Pédol., XI, 97–120 Cornell RA, Posner AM and Quirck J.P (1976) Kinetics and mechanisms of the acid dissolution of goethite (α-Fe OOH). J. Inorg. Nucl. Chem., 38, 563–567 Cornell RM and Schindler PW (1987) Photochemical dissolution of goethite in acid/oxalate solution. Clays and clay Miner., 35, 347–352 Cornell RM, Posner AM and Quirck J.P (1974) Crystal morphology and the dissolution of goethite. J. Inorg. Nucl. Chem., 36, 1937–1946 Cornell R.M., Posner A.M and Quirk JP (1975) The complete dissolution of goethite. J. Appl. Chem. Biotechnol., 25, 701–706 De Endredy AS (1963) Estimation of free iron oxides in soils and clays by a photolytic method. Clay Miner. Bull., 29, 209–217 Deb BC (1950) The estimation of free iron oxides in soils and clays and their removal. J. Soil Sci., 1, 212–220 Duchaufour Ph and Souchier B (1966) Note sur une méthode d’extraction combinée de l’Aluminium et du fer libres dans les sols. Sci. du Sol, 1, 17–29 Farmer VC and Fraser AR (1978) Synthetic imogolite, a tubular hydroxyaluminium silicate. In International Clay Conference., Elsevier, Amsterdam, 547–553 Farmer VC, Fraser AR and Tait JM (1979) Characterization of the chemical structures of natural and synthetic aluminosilicate gels and soils by infrared spectroscopy. Geochim. Cosmochim. Acta, 43, 1417–1420 Follett EAC, McHardy WJ, Mitchell BD and Smith BFL (1965) Chemical dissolution techniques in the study of soil clays. Clay Miner., 6, 23–43 Franzmeier DP, Hajek BF and Simonson C.H (1965) Use of amorphous material to identifiy spodic horizons. Soil Sci. Soc. Am.Proc., 29, 737–743 Hashimoto I and Jackson ML (1960) Rapid dissolution of allophane and kaolinite and halloysite after dehydratation. Clays clay Miner., 7, 102– 113 Henry S (1958) Synthèse de quelques oxydes de fer au laboratoire. C.R. du XXXI Congrès intern. de Chimie Industrielle (Liège)., Mercurius, 1–3 Hetier JM and Jeanroy E (1973) Solubilisation différentielle du fer, de la silice et de l’alumine par le réactif oxalate-dithionite et la soude diluée. Pédologie, 23, 85–99

Selective Dissolution

217

Holmgren GGS (1967) A rapid citrate-dithionite extractable procedure. Soil Sci. Soc. Am. Proc., 31, 210–211 Hsu PH (1977) Aluminium hydroxydes and oxyhydroxyde. In Minerals in Soil Environments, Dixon JB Weed SB and ed., Soil Sci. Sc. Am., 99–143 Hsu PH (1984) Aluminium hydroxides and oxyhydroxides in soils: recent developents. Annu. Meeting and Am. Soc. Agron Jeanroy E and Guillet B (1981) The occurence of suspended ferruginous particles in pyrophosphate extracts of some soil horizons. Geoderma, 26, 95–105 Jeanroy E (1983) Diagnostic des formes du fer dans les pédogénèses tempérées. Evaluation par les réactifs chimiques d’extraction et apports de la spectrométrie Mossbauer. (études des formes organiques du fer amorphe dans les sols)., Thèse Doctorat, Nancy, 109–129 Kampf N and Schwertmann U (1982) The 5M-NaOH concentration treatment for iron oxides in soils. Clays clay Miner., 30, 401–408 Klamt E (1985) Reports of meetings. Iron in soil and clay minerals. Bad Windesheim, West germany, July 1–13 1985. Bull. Soc. Int. Sci. du Sol, 2, 9 Krasnodebska-Ostrega B, Emons H and Golimowski J (2001) Selective leaching of elements associated with Mn-Fe oxides in forest soil, and comparison of two sequential extraction methods. Fresenius J. Anal. Chem., 371, 385–390 Kwong KF and Huang PM (1979) The relative influence of low-molecularweight complexing organic acids on the hydrolysis and precipitation of Aluminium. Soil Sci., 128, 337–342 Lewis DG and Schwertmann U (1979) The influence of Al on iron oxides. Part III - Preparation of Al goethites in M KOH. Clay Miner., 14, 115–126 Lewis DG and Schwertmann U (1979) The influence of Al on the formation of iron oxides. Part IV: The influence of [Al], [OH] and temperature. Clays clay Miner., 27, 195–200 Loveland PJ and Bullock P (1976) Chemical and mineralogical properties of brown podzolic soils in comparison with soils of other groups. J. Soil Sci., 27, 523–540 Loveland PJ and Digby P (1984) The extraction of Fe and Al by 0,1 M pyrophosphate solutions: a comparison of some techniques. J. Soil Sci., 35, 243–250 Mc Keague JA and Day JH (1966) Dithionite and oxalate-extractable Fe and Ag as aids in differentiating various classes of soils. Canad. J. Soil Sci., 46, 13–22 Mc Keague JA (1967) An evaluation of 0,1 M pyrophosphate and pyrophosphate-dithionite in comparison with oxalate as extractants of the accumulation products in podzols and some other soils. Can. J. Soil Sci., 47, 95–99

218

Mineralogical Analysis

Neaman A, Mouélé F, Trolard F, Bourrié G (2004a) Improved methods for selective dissolution of Mn oxides : applications for studying trace element associations. Appl. Geochem., 19, 973–979 Neaman A, Waller B, Mouélé F, Trolard F, Bourrié G (2004b) Improved methods for selective dissolution of manganese oxides from soils and rocks. Eur. J. Soil Sci., 55, 47–54 Norrish K and Taylor RM (1961) The isomorphous replacement of iron by aluminium in soil goethites. J. Soil Sci., 12, 294–306 Petersen L (1976) Podzols and podzolization., Thesis Copenhagen (Danmark) Pollard RJ, Cardile CM, Lewis DG and Brown LJ (1992) Characterization of FeOOH polymorphs and ferrihydrite using low-temperature appliedfield, Mösshauer spectroscopy. Clay Miner., 27, 57–71 Quantin P et Lamouroux M (1974) Adaptation de la méthode cinétique de Ségalen à la détermination des constituants minéraux de sols variés. Cah. ORSTOM, sér. Pédol., XII, 1, 13–46 Quigley RM, Haynes JE, Bohdanowicz A and Gwyn QHJ (1985) Geology, geotechnique, mineralogy and geochemistry of Leda clay from deep Boreholes, Hawkesbury Area. Ontario Geol. Surv., study 29, 128 pages Ryan JN and Gschwend PM (1991) Extraction of iron oxides from sediments using reductive dissolution by titanium (III). Clays and clay Miner., 39, 509–518 Schuppli PA, Ross GJ and McKeague JA (1983) The effective removal of suspended materials from pyrophosphate extracts of soil from tropical and temporate regions. Soil Sci. Soc. Am. J., 47, 1026–1032 Schwertmann U (1964) Differenzierung der Eisenoxide des Bodens durch photochemische extraktion mit saurer Ammoniumoxalate-losung. Z. Pflanzenernähr. Dueng. Bodenk., 105, 194–202 Schwertmann U (1991) Solubility and dissolution of iron oxides. Plant and Soil, 130, 1–25 Ségalen P (1968) Note sur une méthode de détermination des produits amorphes dans certains sols à hydroxydes tropicaux. Cahiers ORSTOM Série Pédol., 6, 106–126 Shuman LM (1982) Separating soil iron and manganese oxyde fractions for microelement analysis. Soil Sci. Soc. Amer. J., 46, 1099–1102 Stol RJ, Van Helden AD and De Bruyn PL (1976) Hydrolysis-precipitation studies of aluminium solution. II - A kinetic study and a model. J. Colloïd Interface Sci., 57, 115–131 Stumm W (1985) The effects of complex-forming ligands on the dissolution of oxides and alumino silicates. In The chemistry of weathering., Reideil D Drever J ed., 55–74 Tamm O (1922) Eine methode zur Bestimmungder anorqanischen komporentem des Gelkomplexes im Boden. Meddal. Statens sSkogförsöksanst, 19, 385–404 Tamm O (1931) Monthly letter., Imperial bureau of soil science, 1 October Tamm O (1934a) Monthly letter., Imperial bureau of Soil Science, 34, August Tamm O (1934b) Über die oxalat-methode in der chemischen Boden analyse. Medd. Skogförsökamsanst, 27 , 1–20

Selective Dissolution

219

Tokashiki Y, Dixon JB and Golden DC (1986) Manganese oxide analysis in soils by combined X-Ray diffraction and selective dissolution methods. Soil Sci. Soc. Amer. J., 50, 1079–1084 Torrance JK, Hedges SW and Bowen LH (1986) Mössbauer spectroscopic study of the iron mineralogy of post-glacial marine clays. Clays clay Miner., 34, 314–322 Yong R, Sethi AJ and La Rochelle P (1979) Significance of amorphous material relative to sensivity in some champlain clays. Canad. Geotechn. J., 16, 511–520

7

Thermal Analysis

7.1 Introduction

7.1.1 Definition Many different analyses of phase transformations involve the use of temperature. This is true in the case of simple gravimetric measurements after drying at a given temperature (cf. Chap. 1), or after chemical precipitation and drying to constant weight. The generic term “thermal analysis” generally only applies to methods carried out according to a dynamically controlled thermal programme making it possible to reveal and quantify different physicochemical transitions. The most common methods used in soil analysis record transformations by means of the temperature either of mass, energy, or the mechanical properties of the samples (Fig. 7.1). Measurements of Mass Variations The abbreviations TGA or TG stand for thermogravimetric analysis and DTG stands for differential thermogravimetry. Losses occur in the form of gases that are simple to detect by (evolved gas detection EGD) and to analyze by (evolved gas analysis EGA). Measurements of Variations in Energy These measurements mainly use DTA, differential thermal analysis and DSC, differential scanning calorimetry. They enable quantification of exothermic or endothermic energy changes at each temperature without inevitably modifying weight.

222

Mineralogical Analysis

Measurements of Dimensional or Physical Variations These include TMA, thermo mechanical analysis (dilatometry) and DMA, dynamic mechanical analysis (viscoelasticity). Though these techniques are used especially in the field of ceramics and plastics, they are also suitable for the study of physical soil properties and specifically for the shrinkage and swelling linked to aggregation and water storage properties (Braudeau 1987, 1988; Braudeau et al. 1999, 2004).

sound, – – –

Fig. 7.1. Table summarizing the principal methods of thermal analysis: TGA, thermogravimetric analysis; DTG, differential thermogravimetry; DTA, differential thermal analysis; DSC, differential scanning calorimetry; TMA, thermomechanical analysis; DMA, dynamic mechanical analysis; STA, simultaneous thermal analysis; EGA, evolved gas analysis. The International Confederation for Thermal Analysis (ICTA) uses the term “derivation” for data resulting from mathematical transformations (derivative calculations on initially measured data) and “differential” for experimental measurements of changes in the delay step concerned

Variations in properties (optical, magnetic, electric, or sound) also occur and can be measured analytically (for example, thermoluminescence is widely used for dating), the loss of magnetic property is represented by the point of Curie, etc. The equipment used for these analyses has reached a high degree of sophistication thanks to its industrial use in the fields of ceramics, glass, cement, plaster, explosives, radioisotopes, pharmaceutics, and polymers where it is widely used for research and quality control. The simplest

Thermal Analysis

223

equipment is not very expensive, but top-of-the-range apparatuses (multiparametric, high temperature, high pressure, etc.) with their associated peripherals and data processing software, can be very costly. 7.1.2 Interest Used in combination with XRD (cf. Chap. 4), IR (cf. Chap. 5) and other chemical analyses, thermal analysis techniques are invaluable tools in mineralogy, geology, pedology, soil chemistry, and physics. They allow the qualitative identification and quantitative analysis of clays (Table 7.1) and many minerals, and also the identification of the different forms of water in soils, oxidation of all forms of organic matter and inorganic materials, phase transitions, etc. The sensitivity of the DTA–DSC methods makes it possible to detect the presence of minerals at the limit of detection of XRD (e.g. goethite, gibbsite at concentration less than 0.25%, substances with short distance atomic arrangement) in clays and soils. All the techniques used enable the balance of the mineral transformations to be established as a function of different geochemical processes both in a weathering profile and in a topographic sequence. Reasonably precise quantitative analysis is possible of hydroxides and oxyhydroxides of iron and aluminium, as well as of clays and particularly of 1:1 kaolinite and halloysite. Continental sediments containing organic complexes can be studied by controlled oxidizing or non-oxidizing pyrolysis, and the evolved gases are analyzed by Fourier transform infrared spectrometry (FTIR). DSC can quantify enthalpy changes during dehydration, dehydroxylation and other forms of structural decomposition in a thermal field extending from –150 to +725°C. The analysis of evolved gases by FTIR or mass spectrometry during temperature scanning makes it possible to calculate the exact chemical transformations undergone by the sample. Thermal analysis is particularly useful for the study of soil genesis, the study of soils rich in para-crystalline compounds (e.g. andosols, histosols) and characterization of the evolution of compounds with shortdistance atomic arrangement that cannot be directly analyzed by XRD. In certain cases it is possible to use the thermogravimetric method in a range of temperatures from ambient to 200°C to indirectly measure the specific surfaces of clays (internal and external surfaces) by impregnation of the sample with a monomolecular layer of an organic material (e.g. the EGME method) or of a gas (e.g. the BET method).

224

Table 7.1. Transformation of clays up to 1,200 °C (adsorbed water and bound water) γ Al2O3 leucite corundum cristobalite ↑ bound water bound water bound waterbound water water bound to water bound to the the crystalline crystalline lattice lattice ≈ 5.1% ≈ 5.0% 700 °C 700°C bound water eustatite....

K

°C

mullite

1,373 1,100 -

metakaolin

1,273 1,000-

end of dehydroxylation

1,073

800- water bound to the water bound crystalline lattice to the crystalline lattice ≈ 5.3% ≈ 4.7%

973

700-

≈ 16.0%

≈ 16.0%

≈ 14.8%

873

600-

≈ 4.8% water in ≈ 4.8% water in cavities cavities ≈ 3.1% 370 °C (350 °C) adsorbed water adsorbed water in cavities adsorbed water ≈ 4.7%

673

400-

Mineralogical Analysis

(350 °C)

(350 °C)

573 (250°C) a (250 °C) (250 °C) (250°C)

300-

Thermal Analysis

(250°C)

in cavities ≈ 0.5%

≈ 3.4%

473 interlayer water ≈ 1.0% water water ≈ 20.0% adsorbed water ≈ 10.1% 2:1 CLAYS micas SMECTITES VERMICULITES Ca
2+

200interlayer water ≈ 20.1% adsorbed interlayer adsorbed interlayer (150°C) (150°C)

adsorbed

adsorbed

adsorbed

adsorbed

adsorbed

water < 0.5%

water

water

water

< 1.0%

≈ 13.2%

≈ 0.5–1.5%

373

100-

≈ 14.0%

1:1 CLAYS

2:1:1 CLAYS

halloysite

kaolinite

4H2 O

pyrophyllite muscovite montmoril- montmoril2+ + lonite Ca lonite Na (Illites)

Na

+

clinochlore

225

226

Mineralogical Analysis

Instrumental thermal dilatometric methods developed by ceramists can be used instead of manual measurements of contraction–dilation (coefficient of linear extension).

7.2 Classical Methods

7.2.1 Thermogravimetric Analysis

Principle Variations in the mass of the sample (losses or increases) are recorded as a function of temperature or time. For soil studies, a temperature range between ambient temperature and 1,100 or 1,200°C is generally satisfactory, but it may be necessary to study reactions up to temperatures of more than 2,000°C, in particular to determine melting points.
Mass variation, % 0

500

1,000

t C

0

500

1,000

t C

Fig. 7.2. Principle of thermogravimetric analysis m = f(t) (on the left) and comparison with differential thermogravimetry dm/dt = f(t) (on the right)

In dynamic analysis, the mass of the sample (m) is heated at a constant rate according to a linear programme based on temperature or time (Fig. 7.2). The extent of the reaction interval, the shape of the curve and the nature of the gases released provide information on the nature of the soil sample and its thermal stability. In TGA by derivation, the first derivative of the variations of mass is recorded as a function of time or of temperature. Two “static” methods can also be used in certain cases, but take a very long time to implement:

Thermal Analysis

227

– The isothermal method where the sample is subjected to a constant temperature and the sample weight is recorded over time until their equilibrium value is reached; – The isobar method where the sample is maintained at a constant pressure and the weight recorded as a function of temperature; the pressure can exceed 500 atm. for certain materials. Clay minerals contain molecules of water and hydroxides bound to the crystal lattice at different energies. During the rise in temperature, this water is gradually moved and eliminated in the form of gas. In the soil, interstitial water of hydration or adsorption that is not bound is generally released first by dehydration. Dehydration does not cause the destruction of the lattice but can cause a modification in the arrangement of the continuous poly- or monomolecular layers (internal and external surfaces of 2:1 clays and interlayer exchangeable ions) along with a contraction of the interlayer space. This phenomenon is reversible. Water in cavities at the base of the tetrahedrons will be more vigorously bound. On the other hand, the phenomenon may be irreversible in soils with allophane, and rehydration will not be able to reach more than around 10% of the initial moisture content. This is also true of hydrated halloysite, 4H2O, which gives metahalloysite, 2H2O. The hydroxyl OH– ions bound to oxygen atoms at the top of the tetrahedral or octahedral units, or present in the external continuous layers of 1:1 clays, or in the internal or external layers of 2:1 clays (or hydrated 1:1 halloysite) are then moved. Their elimination is irreversible and is accompanied by the destruction of the structure (dehydroxylation). The appearance of new forms at a higher temperature can then be recorded, for example kaolinite giving metakaolinite then mullite. But certain transformations that occur without weight loss are not detectable by TGA. In this case the DTA or DSC spectra have to be recorded. It should be noted that the transition between free H2O and OH– of hydroxyls is not always clear as the water bound to the lattice can start to leave before the interstitial water is completely eliminated. In this case thermal analysis at controlled speed of transformation can be used (Rouquerol 1970, 1989; Rouquerol et al. 1985, 1988)1.
1

The method of analysis at controlled speed of transformation makes it possible to measure the initial state of water with more precision. In conventional analysis, the programme of temperature increase varies in a linear way over time. With controlled speed of transformation, the pressure caused by the departure of the gas determines the programme for the increase in temperature by means of a captor. Coupling with a quadripolar mass spectrometer enables analysis of evolved gas. When the adsorbed water with weak activation energy is desorbed, the rise in temperature stops until the initial pressure is restored to the pre-determined point. Then heating continues. This makes it easier to study

228

Mineralogical Analysis

Implementation Reagents – Reagents described in Chaps. 2 and 3 for destruction of organic matter or carbonates, homoionic saturation, etc. – Clays purified for standardization (samples for the reference system of the laboratory and/or international standards). – Saturated solution of magnesium nitrate (Mg(NO3)2, 6H2O, 1,250 g L–1 at 20°C) (to equilibrate the moisture of clays – 56% of relative moisture at 20°C). – Drying agents of varying degrees of effectiveness: (a) Magnesium perchlorate (dehydrite), Mg(ClO4)2, molar weight mw: 223.23; (b) Aluminium oxide Al2O3, mw: 101.94; (c) Silicon dioxide (Silicagel), SiO2, mw: 60.08; (d) Anhydrous calcium chloride, CaCl2, mw; 110.99; (e) Anhydrous calcium sulphate (anhydrite/drierite), CaSO4, mw: 136.04; (f) Phosphorus pentoxide (phosphoric anhydride), P2O5, mw: 141.96. Equipment – Platinum micro-crucibles; – Desiccator with saturated solution of magnesium nitrate; – Desiccators with different drying agents; – Thermal balance (Fig. 7.3). Briefly, a thermal balance is composed of a furnace, a balance and different devices for regulation and data acquisition. Weighing is carried out continuously throughout the thermal cycle. There are two types of equipment: – Balances placed above the mobile furnace; – Balances placed below the mobile furnace. In the first, the sample nacelle is suspended on a wire; in the second, a vertical rod equipped with a support holds the sample. Each system has certain disadvantages that must be minimized to optimize measurements. The balance must allow all losses or increases in mass to be recorded as a function of the temperature and time under all experimental

the mechanisms of hydration and possibly the short- or long-distance atomic structure.

Thermal Analysis

229

conditions. The temperature of the furnace can reach 2,400°C and radiative and magnetic phenomena may occur. The quality of measurements must be the same as with any other analytical microbalance (Pansu et al. 2001). The capacity can vary considerably depending on the use envisaged. For soils, a choice has to be made between (a) relatively big samples, i.e. 100 mg to 1 g with a sensitivity of 10–5–10–6 g and (b) micro-samples, i.e. from 1 to 50 mg (e.g. piezoelectric balances with a sensitivity of 10–8–10–12 g).

Fig. 7.3. Example of a system for thermogravimetric analysis

Different types of balances are available: deflection balances (e.g. with torsion, beam balance, cantilever, with a beryllium bronze spring) but electronic systems without a beam are rarely used because of the disturbances that can be caused by radiative phenomena. The furnace is designed as a function of the temperature. It can be the high frequency induction type or more generally the resistance type (Table 7.2). Kanthal resistances enable measurements at 1,200°C, rhodium resistances at 1,800°C and tungsten resistances at 2,500°C. The furnace is an assembly of metal and ceramic components (high density alumina) which allow resistances to be insulated to ensure

230

Mineralogical Analysis

homogeneous temperatures in the test zone, and also to allow sealing in the case of controlled atmosphere. The furnace has a sophisticated regulation system. The speed of heating must result in a uniform temperature in the test zone and be monitored by thermocouples appropriate for the work temperature (Table 7.3). The programming of the heating cycle must be highly reproducible. The sample must be located in a precise spot in the furnace. The data acquisition system must enable variations in weight and temperature to be recorded simultaneously, and also to carry out a certain number of mathematical operations (e.g. first derivative, surface of the peaks), to control the temperature programme, and finally to store and print the data.

Table 7.2. Type of resistances – thermal ranges material temperature used (°C) melting point (°C) 1,200 1,500 material temperature used (°C) melting point (°C)

constantan nichrome chromel – alumel kanthal platinum (Pt) Pt–Rh 10% Pt–Rh 13% super kanthal

750 1,000 1,100

rhodium (Rh) molybdenuma tantale
b

1,800 2,200 – 2,500 2,850

1,350 1,400 1,755

tungsten

2,800

3,400

1,500 1,700 1,600
a

Hydrogen atmosphere Mechanical resistance up to 1,650°C

b

Procedure The methods have to be standardised if the data obtained in different series is to be compared. The technical characteristics of the thermal balance determine certain obligatory parameters. The nature of the

Thermal Analysis

231

materials to be studied determines the selection of other diagnostic parameters. The samples must be crushed without heating as this could disturb subsequent thermal analysis.
Table 7.3. Types of thermocouples at different temperatures material copper– constantana iron–constantan temperature (°C) 400 (standard) 800 (standard) 1,000 (standard) 1,000 (standard) 1,370 material tungsten – rhenium tungsten – 20% rhenium tungsten tantalum carbide – graphiteb
a

temp. (°C) 2,200

2,400

chromel– constantan chromel – alumel

3,000

Alloy 55% copper + 45% nickel

nickel chromium – nickel platinum–platinum rhodium (10% or 13% Rh)c

b

Argon atmosphere

1,750

c

The thermopiles allow the output signal to be increased without amplification

Initially rough samples (from which organic matter has been eliminated) are reduced by moderate wet crushing. The sample must have a regular particle size and, after air drying, pass through a 0.1 mm sieve. Samples of clays that have been purified and saturated using the methods described in Chap. 3 can also be used. It should be noted that too fine dry grinding can distort the results.2 Adjust the moisture of the sample by placing it in a desiccator containing saturated magnesium nitrate for 48 h (relative moisture should be 56% at 20°C at normal pressure). This treatment homogenizes the hydration layers of any interlayer cations that may be present. Weigh a given weight of sample (5–20 mg) suitable for the range and sensitivity of the balance. Pack the sample in a platinum crucible as

2Dry grinding can modify the nature of the basal faces and consequently their

physical properties. For example, the exchange capacity and some thermal properties of kaolinite can be changed (collapse of the peaks) by too fine grinding which can result in fractures perpendicular to the basal faces and subsequent breakdown of layers.

232

Mineralogical Analysis

regularly as possible to limit differences in thermal diffusity. Place the sample in the thermal balance. Adjust the position of the sample to that of the measurement thermocouple in the furnace. Programme the instrumental variables with the management software, i.e. the linear speed of heating (e.g. 10°C min–1), atmosphere of the furnace, final temperature, etc. Observations For clay the best quantitative measurements are obtained on homoionically saturated samples (Na+, K+, Ca2+ or Mg2+), after elimination of organic matter, soluble salts, ferrous iron, etc. Homoionic saturation of clay enables: – with Na+ ions, improved differentiation between adsorbed water and water bound to the lattice; – with Mg2+ ions improved separation of 2:1 clays from 1:1 clays on the basis of adsorbed water. The presence of organic matter (OM) modifies weight loss of mineral origin. The loss of H2O + CO2 of the OM must be measured. An inert atmosphere can be used to mitigate this phenomenon if it is not too serious, or losses can be estimated by analyzing emitted gas (e.g. by coupled EGA-FTIR). Ferrous iron will oxidize during heating and increase in weight according to the following equations: Fe2+ → Fe3+ + e– or Fe(OH)2 → FeO + H2O↑ 2 FeO + O → Fe2O3 As oxidation of ferrous iron does not result in easily measurable variations in mass, it may be preferable to eliminate it. The same is true for manganese and cobalt. An inert atmosphere can also be used.The elimination of soluble salts avoids secondary recombination. The choice of the nature of the crucible and its geometry can modify the results. The crucible should not react with the sample, with the selected atmosphere or the evolved gas. Certain metals have a catalytic action. Crucibles made of alumina are relatively porous, silver can be used for medium temperatures, platinum can be used for a temperature of 1,500°C. The walls of the crucibles must be as thin as possible (approximately 0.5 mm) to minimize variations in temperature. In the same way, the sample layer should be as thin as possible to ensure that the temperature in the centre is the same as at the edges. As certain minerals are expansible or likely to generate projections, deeper crucibles are sometimes necessary or semi-permeable lids have to be used which, however, can modify the gas flow of the losses.

Thermal Analysis

233

The speed of the increase in temperature influences the decomposition of the sample. For a given temperature interval, slow decomposition is more realistic, than too rapid decomposition which can cause displacement of the characteristic temperatures, a steeper decomposition slope, etc. For exemple, Kotra et al. (1982) showed that a siderite (FeCO3 ) heated at 1°C min–1 had a range of decomposition positioned between 400 and 500°C. At a speed of 20°C min –1 this range moved to between 480 and 610°C. However, a micro sample displays fewer variations than a sample of greater mass. Low speed also enables detection of compounds that only display weak inflection at higher speeds.

Fig. 7.4. Example of TGA and DTG curves for kaolinite (Mâcon, France). Mass of sample: 25 mg; speed of heating: 10°C min –1; atmosphere: nitrogen 30 mL min–1; and platinum crucible.

The atmosphere of the furnace can greatly influence the results with respect to the nature of the decomposition products and types of reactions. Vacuum, inert or reactive atmosphere induce very different thermal spectra. It may be useful to analyze evolved gases.

234

Mineralogical Analysis

Differential thermogravimetric analysis (DTG) In this type of analysis, peaks are obtained whose surface is directly proportional to changes in the mass of the sample (Figs. 7.2 and 7.4), which facilitates quantitative analysis. When the variations in weight are null one obtains a return to the base line for which dm/dt = 0. The curves give an inflection point in the zone where the variation in mass reaches its maximum. The separation of overlapping phenomena is facilitated. Discussion The water contents of a clay can be measured on a sample in equilibrium with an atmosphere at approximately 56% relative moisture (obtained by a mixture of nitrate of magnesium saturated in water). In this way an initial point of reference can be obtained that takes into account complex assemblies of minerals and organominerals, and allows better reproducibility. It is possible to measure water that depends on the exchangeable cations. The water content can be very high for 2:1 clays of the montmorillonite type3 and weak for 1:1 clays of the kaolinite type. However, it is not possible to measure the energy of activation necessary to release water with TGA or DTG. For this purpose DTA must be used giving the quantity of calories that generates the endothermic peak, the quantity of released water being identified by TGA under the same conditions. Existing equipment allows TGA–DTG and DTA to be used simultaneously in the same treatment cycle up to temperatures of 1,750°C or more on the same micro sample. Direct quantification is possible if only one reaction occurs at a given temperature. By insulating the components, quantification becomes possible for organic matter in an oxidizing medium, for carbonates (e.g. dolomite, aragonite, calcite, siderite), sulphides (e.g. pyrites) with very low contents. Here combining quantification with analysis of evolved gas is essential.

3 The layers of 2:1 clays adsorb at least two layers of water molecules. They are

disorientated with respect to one another with a tendency to form association of layers (turbostratic assemblies with 5 or 6 layers).

Thermal Analysis

235

7.2.2 Differential Thermal Analysis and Differential Scanning Calorimetry

Principle of DTA In DTA, the difference in temperature between a soil or clay sample and an inert reference material is recorded as a function of time or temperature with the two substances controlled by the same temperature control programme, at constant linear speed: ∆T = Tsample – Treference. This kind of analysis enables identification of the relations of proportionality that exist between the surface of a peak and the released or absorbed heat during the course of the heating programme. This heat is proportional to the enthalpy of reaction and thus can be used for thermodynamic quantification if the mass of the sample is taken into account. However, in DTA, the simple direct conversion of the peaks into unit of energy is not possible starting only from ∆T as a function of time. Indeed, ∆T depends on the variation in enthalpy, the calorific capacity and the total thermal resistance of the heat flow (R) at a given time. R depends on the nature and the mass of the sample, its preparation (compression, etc.), and the thermal surface of contact between the crucible and the support. These variables are temperature dependent and consequently have to be controlled. For soils, most analysis is carried out between ambient temperature and 1,200°C. When a reaction occurs during an increase in temperature, a difference in temperature is observed between the sample and the reference (Fig. 7.5): – If the temperature is lower than that of the inert reference material, an endothermic peak appears (∆T is negative), this is the case in reactions of dehydration, dehydroxylation, fusion, evaporation, sublimation, etc. – If, on the contrary, the temperature of the sample exceeds that of the reference, an exothermic peak appears (∆T is positive), this is the case for oxidation phenomena (combustion of OM, oxidation of sulphides, oxidation of ferrous iron, certain nucleations or decomposition with neoformation. The recorded differences in temperature are related to the change in enthalpy, but this does not exclude the possibility of two exothermic and endothermic reactions occurring simultaneously. When this is the case, the absence or depression of the DTA peak does not imply the absence of a reaction.

236

Mineralogical Analysis

Contrary to TGA, DTA can produce peaks even if there is no loss or increase in weight, e.g. in reversible second-order transformations: variation in specific heat, magnetic susceptibility, Curie point or α/β allotropic transformation of quartz. TGA and DTA are complementary techniques. The shape, size and temperature of the peaks are influenced by instrumental factors such as the speed of heating, the nature of the sample support and of the thermocouples. Small samples give a better resolution of the peaks and allow faster heating. Slower speed can increase sensitivity, but to the detriment of temperature, precision, and resolution. A dynamic atmosphere is preferable to a self-generated static atmosphere. This allows continuous evacuation of evolved gas, thus reducing the risks of artefact reactions at higher temperatures. These gases can then be analyzed enabling identification of the molecular structure of the compounds that caused the gaseous emission.

Fig. 7.5. Diagrammatic output of differential thermal analysis. BD, peak width; EC, peak height; BCD, peak surface; AF, base line

T T T –
endothermic peak

T

The range of different types of equipment sold by different manufacturers means complex mathematical demonstrations are not required to validate the different parameters taken into account, see for example Duval (1963), Watson et al. (1964), Garrels and Christ (1965), Allen (1966), Gray (1968), Mackenzie (1970), Brennan (1971), Miller et al. (1973), and McNaughton and Mortimer (1975). Principle of Differential Scanning Calorimetry Other techniques to measure energy are grouped under the name of DSC. DSC techniques are often badly defined because patents use the same term for different concepts.

Thermal Analysis

237

The term DSC applies to apparatuses able to measure specific heat, or the heat capacity of a sample, and to quantify the energy of the reactions during the heat treatment. In DSC with power compensation, the sample and reference are continuously maintained at the same temperature by individual resistances. The parameter that is recorded is the quantity of power consumed by the compensation resistances, that is to say d(∆Q)/dt or dH/dt in millicalories per second as a function of the temperature (controlled linear increase). When a reaction occurs in the sample, thermal energy is added or removed. The quantity of energy added or removed is equivalent to the quantity emitted or absorbed by a given transition. The recording of this balance of energy is a calorimetric measurement of enthalpy.
T Tsample -Treference T Tsample-Tref

Sample ref

T

Fig. 7.6. Thermal systems using measurements of energy: a, classical DTA; b, Boersma DTA with heat flow or conventional indirect DSC; and c, DSC with compensation of power

In the Boersma technique, also known as DSC or DTA with heat flow, ∆T = Tsample – Treference are measured like in DTA. The sensors are placed below the crucibles to reduce the effects of variations in the thermal resistance of the sample. This way of assembling the components is

238

Mineralogical Analysis

similar to other DSC heat flow equipment that records d∆q/dt type data or, with thermopiles, dQsample/dt – dQreference/dt = d(∆Q)/dt enabling the quantitative measurement of enthalpy. In modulated DSC, the sample is subjected to a linear increase in heating (for example 10°C min –1) on which a sinusoidal modulation of temperature (of 30 s and amplitude of 1°C) is superimposed resulting in a cyclic heating profile. There is thus a speed of constant subjacent heating and instantaneous measurements at sinusoidal speed. The deconvolution of the flow profile resulting from the heating profile provides the total heat flow as in conventional DSC, but also separates flow into two components: reversible specific heat and non-reversible kinetic heat. This technique enables improvement of resolution and sensitivity, because low speed favours good resolution and high heating speed favours good sensitivity. The functioning temperature of the DSC apparatuses with power compensation is generally limited to 750°C, which is too low to study certain reactions in the soil. Classical DSC makes it possible to reach temperatures similar to those of DTA or TGA. Materials These must be reference products and be ground to 0.1 mm in an agate mortar. – Products of known melting points to calibrate temperature (Table 7.4). – Inert reference: alumina (Al2O3) calcined to 1,200°C, and ground to 0.1 mm in an agate mortar does not show any effect of heating except possibly a few irregularities in the base line. – Purified clays belonging to the laboratory reference set homoionically saturated by Mg2+, Ca2+, or Na+. Clay samples from industrially exploited sedimentary deposits should be distinguished from those coming from soils. – Pure minerals from the laboratory reference set and industrial reference products for analysis. – 1,250 g (Mg(NO3)2,6H2O) L–1 of magnesium nitrate saturated aqueous solution at 20°C. Equipment – DTA micro crucibles made of platinum (or possibly alumina, quartz, tungsten, zirconium oxide, nickel, or aluminium). – Desiccator with Mg(NO3)2, 6H2O saturated solution. – Desiccator with desiccating agents.

Thermal Analysis

239

– Analytical balance (10–6–10–8 g) with a weighing range of 1–100 mg for quantitative analysis. – Apparatuses: industrial DTA–DSC instruments are designed and built according to different concepts, and this can make comparison of data difficult. It is advisable to use a laboratory reference set to ensure the precision and relevance of the results. The equipment generally comprises different components (Figs. 7.6 and 7.7):
Table 7.4. Calibration products for thermal analysis product naphthalene phenanthrene formula melting point (°C) 80.2 mw (g mol–1) 128.16

C10H8 C14H10

99.3
COOH

178.22 122.12 244.31 172.18 118.70 172.42 65.38 104.87 60.06 26.98 100.09 107.88

benzoic acid barium chloride
a

C7H6O2 BaCl2,2H2O CaSO4,2H2O Sn CdCO3 Zn

122.3 (sublimation) 130.0 (–2 H2O) 180.0 (–2 H2O) 231.97 350.00 419.4 (ignition at air contact) 553 573 660.2 850 (CO2 + CaO) 960.8

calcium sulphate

tin cadmium carbonate zinc lithium bromide quartz aluminium
a

LiBr,H2O (deliquescent) SiO2 Al CaCO3 Ag

calcium carbonate

(468 cal g–1) silver
a

Standard material used to measure heat of reaction.

– A furnace equipped with a device to programme temperature, to control atmosphere and to accelerate cooling (between 20 and 40 min before a new cycle starts); – A support for the sample equipped with differential temperature detectors; – A management station for analytical programmes and recording of measurements.

240

Mineralogical Analysis

Temperatures exceeding 2,000°C are required for ceramics. For soils, a temperature of 1,200°C is generally sufficient but 1,600°C may be needed. The position of the sample support and the inert reference can vary considerably (Fig. 7.7). The thermocouples can be placed in or outside the crucibles (or to even welded into a cavity in the crucible). Metal crucibles give smaller peaks than ceramics with a faster heat transfer thus limiting the risks of deformation of the peaks. For studies up to approximately 1,000°C, chromel–alumel thermocouples generate a significant electromotive force (EMF = 45.16 mV at 1,100°C) but relatively low chemical resistance. On the other hand, platinum–rhodium or platinum thermocouples generate EMF = 10.74 mV which requires amplification, but can reach temperatures of 1,500°C and are chemically much more resistant.
Reference Sample Regulation

Power Thermocouple Oven Fig. 7.7. DTA equipment

Data acquisition and control

T Amplifier T (C)

The position of the thermocouples in the sample and in the reference must be perfectly symmetrical to avoid any variation in the shape of the peaks or displacement of the base line. The measurement station should include a micro computer and associated peripherals and ensure: – Complete piloting of the temperature cycles (speed of heating, 1–50°C min–1 with stabilized voltage, temperature control by means of thermocouples, carrying out instructions, etc.; – Control of the atmosphere (purging circuits, admission of carrier gas, admission of inert or reactive gases, EGA output, regulation of flows and pressure);

Thermal Analysis

241

– Simultaneous acquisition of data on temperature and heat transfers with very weak inertia, thermodynamic calculations and print-outs of the results. As is true for TGA, standardization of the procedures is essential. Sample Use a sample of whole soil ground to 0.1 mm (should not be overground). Organic matter in humic soils produces a strong exothermic reaction culminating at 300°C (573 K) which obstructs the endothermic peaks of dehydration towards 250–400°C if analysis is carried out in air or in an oxidizing atmosphere. It is consequently better to destroy OM using the procedures described in Chaps. 2 and 3 (or to extract it and use DTA). On whole soil, the peaks are generally of low intensity. Quartz can obstruct, but its peak at 573°C can be used as a marker. To improve sensitivity, it may be better to use isolated purified and enriched fractions of a given particle size, homoionically saturated in the case of clays and, after air drying, stored in a solution saturated with Mg(NO3)2,6H2O to maintain constant relative moisture (the size and crystallinity of the particles influence the temperature, size, and shape of the peaks). In sandy fractions (2.0–0.05 mm) primary minerals and concretions rich in iron and manganese can be measured. The endothermic peak of quartz is generally observed at 573°C. Silt fractions (0.05–0.005 mm) often represent a more complex medium: they contain both primary minerals and deteriorated forms and possibly forms of clay minerals > 2 µm. Coarse or fine clays (< 2 µm) often give much more intense and quantifiable peaks than whole soils. Substances with short-distance arrangement (e.g. allophane, ferrihydrites) are found in this fraction. Homoionic saturation by a cation (Na+, K+, Ca2+, Mg2+, Al3+) enables certain properties such as levels of hydration to be revealed. The first endothermic reaction of volatilization of the adsorbed water varies with the number of water molecules associated with the cation. The cations affect the size and the shape of the peaks, generally without significantly modifying the temperature at which they appear. It is possible to mix the clay fraction with alumina to avoid caking, retraction, and cracking of the sample, but this has a diluting effect.

242

Mineralogical Analysis

Two clay samples from the same profile can be compared under the same experimental conditions. The reference sample will be then one of the samples. If they are identical, ∆T will be equal to 0 and there will be a base line with no peak, as the two samples will cause the same phenomenon to occur at the same time. Reference Material The standard must be thermically inert in the temperature range to be used. It must have a thermal conductivity or a thermodiffusivity close to that of the sample. The particle size should be that of soil (ground to 0.1 mm) which allows close contact with the crucibles as well as a favourable bulk density (same proportion of pores near the ground soil). Generally alumina is used (Al2O3 calcined at 1,200°C for 1 h). This reference can be used for several measurements, but slight hygroscopicity of the product may be observed after 4 or 5 measurements. A clay burnt at 1,200°C and then ground can also be used, but problems often occur with respect to the base line. For example between rough kaolinite and kaolinite calcined to mullite at 1,200°C, thermal conductivity will vary by a factor of two and certain reversible reactions may occur. Quartz, magnesia, or aluminium sheets are sometimes used as reference materials. Procedure The procedure is similar to classical qualitative or quantitative DTA– DSC analysis, and takes into account the factors mentioned earlier. – Weigh on the laboratory balance (± 1/100 mg) a given mass of sample (1–100 mg), suited to the characteristics of the apparatus and an identical quantity of inert reference; for DSC with power compensation, an empty crucible similar to that used for the sample can be used as inert reference. – Pack as homogeneously as possible in platinum crucibles. – Find the optimal position for the crucibles (in suitable supports) with respect to the thermocouples and the most homogeneous zone of the furnace.

Thermal Analysis

243

– Cover the crucibles with a lid if required (the displacement of evolved gas will be modified). – Regulate the speed of linear increase in temperature; higher speed can increase some peak temperatures, but can also improve sensitivity (peak height); generally the range is between 5 and 20°C min–1 but certain materials allow speeds of 0.1–200°C min –1 (it is also possible to use the reverse technique with recording during cooling). – Adjust the atmosphere of the furnace taking into account possible analysis of the gaseous phase by interfacing with EGA. Software packages can monitor and control operating conditions including calculations and printing out the results, in this case all the parameters are optimized. Interpretation of the Results
Table 7.5. Main thermal effects on the clays of soils range of temperature (°C) 50–250

signals

endothermic peaks loss of absorbed water, interlayer water (e.g. allophanes, halloysite, smectite) exothermic and endothermic peaks crystallisation reactions (e.g. gels of iron or aluminium oxides) diagnostic peaks of the crystalline forms reactions of dehydroxylation exothermic peaks reaction of nucleation or recrystallization (e.g. kaolinitechlorites)

400–700

900–1050

The same procedure is used for rough samples and extracted fractions of different degrees of purification under inert or reactive atmosphere depending on the nature of the sample studied. “Pure” standard minerals or minerals in known mixtures (90:10, 80:20%, etc.) can be treated to trace the calibration lines with the same criteria thus enabling quantification of the samples (a reference set of laboratory minerals is necessary). A distinction must be made between clay minerals from sedimentary deposits whose peaks are generally well defined and minerals from different soils after purification and concentration.

244

Mineralogical Analysis

First, well-defined peaks are identified (e.g. the endothermic peak at 500°C and the exothermic peak towards 900°C for kaolinite). The shape of the peaks is an indicator of the thermal process: fusion results in a narrow endothermic peak, decomposition often produce a broad endothermic peak which is often asymmetrical, a second-order transition results in only a slight inflection. Certain minerals give well-defined peaks that are used as markers of characteristic transitions in the thermal programme. The slope of the base line and the magnitude of the peaks should be taken into account. The main thermal transitions observed in clays are listed in Table 7.5. The “low temperature” zone provides information on adsorbed water and on the presence of certain clays and amorphous substances. The medium temperature zone shows peaks resulting from the predominant influence of crystallinity. Finally the high temperature zone shows phenomena such as nucleation or recrystallization of clays. Certain standards are used in quantitative analysis like CaSO4,2H2O to enable calculation in millicalories per unit of area (calibration between a peak surface and a known emitted or absorbed heat). Each type of structure can be identified by a characteristic thermal curve (Fig. 7.8). For clays, purification is necessary to limit the artifacts caused by impurities which are likely to be superimposed in the same temperature zone. Clays with a tetrahedral layer and an octahedral layer will display phenomena related to the decomposition of the octahedral layer. The exothermic volatilization of hygroscopic water may be accompanied by a recombination of the elements of the tetrahedral and octahedral layers. Quantitative DTA is possible for clays presenting endothermic peaks associated with loss of water bound to the crystal. This is the case of 1:1 clays like kaolinite, but 2:1 clays like montmorillonite, vermiculite, or illite are difficult to quantify with this method. Deferrification reinforces the exothermic peaks of kaolinite, and iron inhibits exothermic phenomena. The presence of carbonate, organic matter, and alkaline ions must be taken into account. Hydroxides initially present loss of water, and then contraction linked to decomposition of the lattice. The heights of peaks are faster to use than the surfaces of peaks and can thus be used for approximations. Thermodynamic calculations use all the spectral information as well as the quantities of evolved water measured in TGA.

Thermal Analysis

245

Fig. 7.8. Examples of DTA thermal diagrams on soil minerals (J Gautheyrou, set of reference standards from IRD Bondy, France, unpublished data)

In DSC, the curves are proportional to the enthalpy: peak surface = m ∆H/k ∆H; heat of transition; m; reactive mass; k; coefficient of calibration (independent of the temperature measured by DSC with power compensation).

246

Mineralogical Analysis

7.3 Multi-component Apparatuses for Thermal Analysis

7.3.1 Concepts

Fig. 7.9. Example of gas sweeping devices for TGA or TGA–DTA. (a) Simple displacement by a neutral carrier gas crossing the balance and protecting it from corrosion; (b) system with two gases – one neutral and one reactive – with horizontal sweeping of the sample; and (c) system with purging by vacuum and later sweeping by neutral gas or neutral gas + reactive gas

The development of new apparatuses was based on several different concepts. In multi-component equipment, several dedicated apparatuses are run by a management station and measurement is computerized. This type of multi-function equipment enables multiple measurements on the same sample subjected to a thermal profile with controlled atmosphere and pressure. Suitable sensors can measure TGA, DTG, DTA and DSC simultaneously thus reducing the time required for analysis, increasing precision and allowing more thorough characterization. Automated introduction of the samples in the furnace is useful in the case of equipment with a cooling cycle of less than 30 min (return to ambient temperature after heating to 1,700°C).

Thermal Analysis

247

Certain dedicated measurement peripherals can also detect or quantify evolved gas. This quantification is essential to complete certain DTA spectra (e.g. simultaneous exothermic and endothermic reactions). The methods selected can be programmed and used in sequence, providing a powerful reference frame, e.g. for the identification of clays and the study of transformations in structure, recrystallization, forms of water, thermal stability or thermal oxidation. 7.3.2 Coupling Thermal Analysis and Evolved Gas Analysis The analysis of emitted or emanating gas may require coupling of equipment of varying degrees of complexity. A small furnace with an open configuration should be used that is able to ensure rapid evacuation of decomposition gases (Fig. 7.9). The circulation of a carrier gas and the possible presence of reactive gases causes changes in pressure and convection phenomena linked to pressure and local temperatures (as well as thermomolecular forces able to create heterogeneous temperatures inside the furnace) that are controlled by a system of screens. Below is an overview of the most widely used systems. Simple detection of evolved gas (EGD): a neutral carrier gas draws evolved gas into a catharometer detector; the output signal is proportional to the concentration of evolved gas in the carrier gas. The nature of the evolved gases is not identified. It is possible to trap the heavy products in liquid nitrogen, and then to subject these products to further analysis, for example controlled pyrolisis by thermal analysis. Analysis of evolved gas (EGA) can be performed by coupling equipment of varying complexity. Gas chromatography, coupled with simultaneous TGA-DTA equipment, makes it possible to characterize evolved gas by automatic discontinuous injections (it is also possible to trap the fractions that are not analyzed in liquid nitrogen). Integration of the peaks makes it possible to quantify evolved and separated gas as a function of their retention times. Coupling a Fourier transform infra-red spectrometer (FTIR) enables analysis of time of flight thanks to the speed of acquisition and the sensitivity of FTIR. A flow of nitrogen automatically transfers the evolved gases towards the spectrometer. Resolution is about 4 cm–1. Each temperature point is stored in a numerical file corresponding to the number of the selected wavelength at this point. Water is identified in the 3,600 and 1,600 cm–1 zones, CO and CO2 in the 2,000 to 2,400 cm–1 zones.

248

Mineralogical Analysis

Fig. 7.10. Decomposition of the standard calcium oxalate (CaC2 O4 , H 2O) studied by coupling TGA, DTG and DTA (at the top) and mass spectrometry (at the bottom) atmosphere, air; speed of heating, 5°C min–1, sample mass, 50 mg. 1. CaC2O4,H2O → CaC2O4 + H2O↑ (12.2% H2O) → CaCO3 + CO↑ (19.1% CO) 2. CaC2O4 → CaO + CO2↑ (30.1% CO2) 3. CaCO3

Thermal Analysis

249

Coupling with a mass spectrometer is very powerful, but difficult to implement, as the mass spectrometer operates under a vacuum of about 10–5–10–6 mbar. Interfacing with the thermal analyzer is achieved by means of double stage separators. A quadripole mass spectrometer enables rapid scanning in the mass range of evolved gas up to approximately 100, including 12C, 16CH4, 28CO, 44CO2, 18HOH, 16O, 32O2, 34 H2S, 64SO2, 80SO3. This range is not sufficient for the observation of organic materials resulting from controlled pyrolysis (cf. Chap. 12) as the molecular weights are distributed over a larger range, up to 500 or more. Figure 7.10 shows analysis of the thermal decomposition of oxalate of calcium, which can be used as a standard for calibration. With a combined mass spectrometer with a relatively low speed of acquisition and air atmosphere, decomposition reaction 2, which gives calcium carbonate and carbon monoxide, gives only one peak at mass 44 (CO2). With nitrogen atmosphere, formation of a CO–CO2 mixture is observed, showing the beginning of decomposition of the newly formed carbonate. Simultaneous detection of CO and CO2 can also be achieved by coupling the thermal analyzer with a Fourier transform infra-red spectrometer. The analysis of radioactive gases emanating from rocks in their solid state at different temperatures can be performed by coupling thermal analysis with a detector of α-particles.

References
Allen JA (1966) Energy Changes in Chemistry., Allyn-Bacon Newton, MA Braudeau E (1987) Mesure automatique de la rétraction d’échantillons de sol non remaniés. Science du Sol, 25, 85–93 Braudeau E (1988) Equation généralisée des courbes de retrait d’échantillons de sols structurés. Comptes rendus acad. Sci. Fr., série 2, 307, 1731–1734 Braudeau E, Costantini JM, Bellier G and Colleuille H (1999) New device and method for soil shrinkage curve measurement and characterization. Soil Sci. Soc. Am. J., 63, 525–535 Braudeau E, Frangi JP and Mothar RH (2004) Characterizing non-rigid dual porosity structured soil medium using its characteristic shrinkage curve. Soil Sci. Soc. Am. J., 68, 359–370

250

Mineralogical Analysis

Duval C (1963) Inorganic Thermogravimetric Analysis., Elsevier. Amsterdam. 2ème édition Garrels RM and Christ CL (1965) Solutions, Minerals and Equilibra., HarperRow New York Gray AP (1968) Symposium Analytical Chlorimetry, Porter R.S. and Johnson JF ed., Proc. Am. Chem. Soc., Plenum New York, 209 Kotra RK, Gibson EK and Urbancic MA (1982) Icarus., 51, 593 Mackensie RC (1970) Differential Thermal Analysis. Academic London. Vol. 1. McNaughton JL and Mortimer CT (1975) La calorimétrie differentielle à balayage. Butterworth, Physical Chemistry., Série 2, 10. 43 pages Miller B, Giuham JM, Brennan WP, Nentzer C and Whitwell JC (1973) Thermochim. Acta, 5, 257 Pansu M, Gautheyrou J and Loyer JY (2001) – Soil analysis – sampling, Instrumentation and Quality Control, Balkema Amsterdam, Lisse, Abington, Exton, Tokyo, 512 p Rouquerol J (1970) L’analyse thermique de décomposition construite. J. Therm. Anal., 2, 123–140 Rouquerol J (1989) Reciprocal thermal analysis. The hidden face of the thermal analysis. Thermodyn. Acta Rouquerol F, Rouquerol J, Thevand G and Triaca M. (1985) Desorption of chemisorbed species: its study by controlled rate thermal analysis. Surf. Sci., 239–244 Rouquerol J, Rouquerol F, Grillet Y and Ward RJ (1988) A critical assessment of quasi equilibrium as adsorption techniques in volumetry gravimetry, calorimetry. In Characterization of Porous Solid, Kunger KK ed., Elsevier Amsterdam, 67–75 Watson ES, O’neill MJ, Justin J and Brenner N (1964) DSC. Anal. Chem., 36, 1233

Bibliography
Bang DV and Atanasov I (1984) Thermal characteristics of the clay fraction of soils overlying zeolites. Pochvoznanie i Agrokhimiya, 19, 58–65 Barshad I (1965) Thermal analysis techniques for mineral identification and mineralogical composition. In Methods of Soil Analysis, Part 1., Black C.A. ed., A.S.A., S.S.S.A., 9, 699–742 Bishop JL, Banin A, Mancinelli R and Klovstad MR (2002) Detection of soluble and fixed NH4+ in clay minerals by DTA and IR reflectance spectroscopy: a potential tool for planetary surface exploration. Planetary Space Sci., 50, 11–19 Brennan WP (1974) Application of differential scanning calorimetry for the study of phase transitions. In Analytical Calorimetry, Porter RS and Johnson JF ed., Plenum New York Caillere S and Henin S (1963) Mineralogie des argiles., Masson Paris Daniels T (1973) Thermal Analysis., Kogan Page

Thermal Analysis

251

Dunn JG (1980) L’analyse thermique, une technique de centrale de qualité dans les industries de l’argile, des céramiques et des verres. Silicates Industriels, 10, 203 Duval C (1953) Inorganic Thermogravimetric Analysis., Elsevier Amsterdam. 1st edition Earnest CM (1983) Thermal analysis of Hectorite. Part II. Differential thermal analysis. Thermochim. Acta., 63, 291–306 Emmerich WD and Kaisers Berger E (1979) Simultaneous TG-DTA massspectrometry to 1550°C. J. Therm. Anal., 17, 197–212 Ferenc Paulik (1995) Special Trends in Thermal Analysis., Wiley New York, 478 p Fordham CJ and Smalley IJ (1983) High resolution derivative thermogravimetry of sensitive clays. Clay Sci., 6, 73–79 Gallagher PK (Ed.) (1998) Handbook of Thermal Analysis and Calorimetry., Elsevier Garn PD (1965) Thermo Analytical Methods of Investigation., Academic London Giovannini G and Lucchesi S (1984) Differential thermal analysis and infra-red investigations on soil hydrophobic substances. Soil Sci., 137, 457–463 Hatakeyama T and Zhenhai Lui (Ed.), (1998) Handbook of Thermal Analysis., Wiley New York Keyser de WL (1953) Differential thermobalance: a new research tool. Nature, 172, 364 Khanafer K and Vafai K (2002) Thermal analysis of buried land mines over a diurnal cycle. IEEE Trans. Geosci. Remote Sensing, 40, 461–473 Lombardi G (1984) Thermal analysis in the investigation of zeolitized and alterad volcanics of Latium, Italy. Clay Miner., 19, 789–801 Mackenzie RC (1963) SCIFAX, Differential Thermal Analysis Data Index., Cleaver-Hume Press Mackenzie RC, Keattoh CJ, Dollimore D, Forester JA, Hodgson AA and Redfern JP (1972) Nomenclature in thermal analysis II. Talanta, 19, 1079–1081 Mackenzie RC and Caillere S (1975) The thermal characteristics of soil minerals and the use of these characteristics in the qualitative and quantitative determination of clay minerals in soils. In Inorganic componants, Gieseking E. ed. vol. 2, Springer Berlin 529–571 Madkensie RC and Robertson HS (1961) The quantitative determination of halloysite, goethite and gibbsite. Acta Univ. Caral. Geol. Suppl., 1, 139 Maizenberg MC, Karpachevski LO, Markovich MN and Kurakof VN (1991) The use of differential scanning calorimetry in soil studies. Moscow Univ. Soil Sci. Bull., 46, 31–34 Muller F, Drits V, Plancon A and Robert JLL (2000) Structural transformation of 2:1 dioctahedral layer silicates during dehydroxylation– rehydroxylation reactions. Clays Clay Miner., 48, 572–585 Murphy CB (1962) Differential thermal analysis. A review of fundamental developments in analysis. Anal. Chem., 298–301 Paulik F, Paulik J and Arnold M (1984) Simultaneous TG, DTG, DTA and EGA technique for the determination of carbonate, sulphate, pyrite and organic materials in minerals, soils and rocks. II – Operation of the

252

Mineralogical Analysis

thermo-gaz-titrimetric device (TGT) and examination procedure. J. Therm. Anal., 29, 333–344 Poerschmann J, Görecki T and Parsi Z (2005) Analytical non-discriminating pyrolysis in soil analysis. LabPlus Int., 19, 8–14 Sarikaya Y, Onal M, Baran B and Alemdaroglu T (2000) The effect of thermal treatment on some of the physicochemical properties of a bentonite. Clays Clay Miner., 48, 557–562 Schnitzer M and Kodama H (1972) Differential thermal analysis of metal-fulvic acid salts and complexes. Geoderma, 7, 93–103 Schomburg J (1991) Thermal reactions of clay minerals: their significance as “erchaeological thermometers” in ancient potteries. Appl. Clay Sci., 6, 215–220 Schultze D (1969) Differential Thermo-Analysis., Verlag Smothers WJ and Chiang YZO (1966) Handbook of Differential Thermal Analysis., Chem. Publ. Co. Smykatz-Kloss W and Heide K (1988) Progress of thermal analysis in earth Sciences. J. Therm. Anal., (GBR) 33, 1253–1257 Taichev T, Konishev P and Donov D (1984) Application of thermal analysis in studies on the structural evolution of humic substances. Pochvoznanie i Agrokhimiya, 19, 82–87 Tan KH and Clark FE (1969) Polysaccharide constituents in fulvic and humic acids extracted from soil. Geoderma, 2, 245–255 Tan KH, Hajek BF and Barshad I (1986) Thermal analysis techniques. In Methods of soil analysis, Klute A. ed., A. S..A., S.S.S.A., 9, 151–183 Trofimov SA, Tolpeshta II, Sokolova TA (1999) A study of plant material and organic soil horizons using thermal analysis techniques. Pochvovedenie, 54, 3–9 Van Olphen H and Fripiat JJ (1979) Data Handbook for Clay Minerals and Other Non-Metallic Minerals., Pergamon New York, 350 pages Wendlandt WW (1974) Thermal Methods of Analysis, Part 1. Wiley New York

8 Microscopic Analysis

8.1 Introduction
The study of the processes of genesis and weathering of the soils and the characterization of minerals require observations at different spatial and temporal scales. Pedon, horizon, macro- and micro-samples are used as a basis for a range of examinations in which microscopy techniques enable observation of the interfaces of the different phases and even of the ultimate stages of molecular assemblies. Complementary analytical probes enable in particular determination of the chemical nature (e.g. energy dispersive X-ray, EDX or wavelength dispersive X-ray, WDX analysis) and the structural organization (electron micro-diffraction) of these phases. Whereas classical wet chemical analyses provide information on overall evolution, microscope analyses enable identification of hyperfine chemical heterogeneity and facilitate understanding of transformation mechanisms. Qualitative and quantitative information is obtained on the texture of individual particles, e.g. shape, the presence of cements, inclusions, pores. Detailed morphological analysis (spatial relations of the different phases, preferential orientation, etc.) is performed by coupling a microscope to an image or texture analyser, and analysing thin sections to reveal the cartography of chemical segregation, micro-profiles, and to improve our knowledge of pedogenesis. The crystalline structure can be identified by changing scale: i.e. by studying the elementary mesh of clays, regular sequences of silicates in layers, modifications of the basal interlayer, the appearance of aperiodisms, defects, micro-cleavages, growth steps, gels with shortdistance atomic arrangements near source crystals which bring about the transformation of one mineral into another at the nano-structural scale. Electron micro-diffraction, which is available on the majority of electron

254

Mineralogical Analysis

microscopes, enables identification of crystalline structures. All these options make microscopy an extremely powerful tool. Optic and electronic microscopy and the accompanying peripherals used to determine chemical distribution within the sample are widely used in pedology, mineralogy, crystallography, petrography, metallography, and clay geology, and provide precise information on the mechanical properties, localization of surface charges and exchange properties of the soils. However, it is extremely important to think about the validity of the observations and to adapt the scale of measurement to the exact requirements of the sample. Sampling conditions and the different stages of sample preparation should always be specified. The season should also be taken into account (samples taken in a wet or dry season present very different pore spaces in soils rich in 2:1 clay). Not only pedogenesis and geostatistics but also geomorphology must be used (the samples may require complementary sampling of the evolution of a profile, or sampling at a larger scale to account for the homogeneity of a zone). When the observations are made, it is important to be aware of the physical and chemical processes that can modify a sample and lead to erroneous results (e.g. contamination, oxidation, neoformation). It is also important to explore the full potential of the equipment, as the full capacity of apparatuses is frequently underexploited. Electron beams can damage the surface of materials. The depth of penetration of radiation may be limited to near the surface: i.e. approximately 1–3 nm, or on the contrary, may gradually erode the surface resulting in a concentration of profiles (e.g. in plasma bombardment, secondary ion mass spectrometry). As is true for all methods, the calibration of measurements is essential and requires a reference set of sample observations that will allow comparison with known and clearly identified situations.

8.2 Preparation of the Samples

8.2.1 Interest Visual examination, followed by examination with a magnifying glass is not enough as it only enables a rough estimate of modes of assembly and physical properties. This type of examination is consequently always supplemented by tactile examinations and other sensorial and chemical tests (described in Pansu et al. 2001).

Microscopic Analysis

255

Observations can be made under an optical or electron microscope on soil samples that have undergone treatments such as: – purification, concentration, quantitative separation into classes of particle size – separation in heavy liquids of increasing densities, up to d = 4.28 at 20°C (e.g. bromoforme, tetrabromomethane, di-iodomethane, Clerici solution) – magnetic separation in a Frantz magnetic separator (or similar). In this case, individual particles are measured (shape, nature, relative proportions of minerals, effects and nature of weathering). To obtain precise information on assemblies, thin sections have to be prepared on glass slides making it possible to study soil transformation and to link this information with field observations. 28 × 48 mm petrographic slides are the most widely used, but 110 × 76 mm or even 200 × 180 mm Mammoth slides are also useful for micropedological and micro-morphological studies (however, these are very delicate operations). 8.2.2 Coating and Impregnation, Thin Sections

Principle The term coating is generally used in the case of massive compact samples with simple geometry, such as pieces of hard stone which can be embedded in a resin that hardens rapidly and facilitates preparation for analysis. The purpose of impregnation with resin is to consolidate soft porous rocks, products of rock deterioration and soils (Delvigne 1998) for which cohesion is required. Wet soils must be dried in a way that avoids modifying the structure of the sample by contraction or the appearance of cracks due to shrinkage. One way to limit this phenomenon in soils rich in 2:1 clays with low porosity is to saturate the sample with sodium chloride before desiccation by freeze-drying. For soils rich in allophanes with a water storage capacity reaching 250% of dry soil, water can gradually be replaced by acetone, or dioxane. This technique is also applicable on clay soils, but in this case, samples treated with acetone are more fragile than those treated with dioxane (Tessier 1985). Acetone and dioxane are compatible with certain resins used for impregnation. Freeze-drying generally preserves the features of the sample better than oven or air drying. For small samples (approximately 30 cm3) the

256

Mineralogical Analysis

use of a dehydration apparatus at CO2 critical point gives excellent results. For bulky wet samples, the use of epoxy resins that are soluble in water is one possible solution (Moran et al. 1989). Preparing the thin section slides manually is very time consuming and considerable technical skill is needed to obtain thin sections of quality with the required degree of transparency and a uniform thickness of between 20 and 30 µm. Automation of coating and impregnation, slicing, and polishing is possible, but requires a high initial investment which can only be amortized by at least 60 coatings or 40 impregnations per week, i.e. approximately 2,000 slides of format 28 × 48 mm per year. Some computer-controlled modular systems enable automated production of super-thin sections 10 µm thick, with 0.25 µm polishing if required. Standard slides are 30 µm thick with approximately 1 µm polishing. (Fig. 8.1) It is often preferable to have the thin section slides made in specially equipped laboratories because of the need for explosion-proof electric circuits, vapour evacuation circuits with solvent traps, plus the cost of maintenance of the equipment and the instability of the resins. Equipment – Laboratory equipped with a fume hood (with non-deflagrating electric apparatuses) – combined system for grinding and polishing with a disc 250 mm in diameter – cutting saw with a diamond grinding stone 350 mm in diameter – equipment for cold impregnation and coating under vacuum: desiccator 300 mm in diameter with a vacuum stopcock and a device for admission of the mixture of resin + catalyst – vacuum pump with pressure gauge (0–750 mm Hg) – ultrasound tank – drying oven with ventilation (up to 100°C) – stylograver with a tungsten carbide point – petrographic glass slides, size 28 × 48 mm, 1.2 mm thickness – glass sheets 0.13 mm thick for use as slide covers – boxes for thin section slides – ultramicrotome with thermally regulated system of displacement – freeze dryer – vacuum desiccators with drying products such as silica gel – binocular microscope – small equipment such as glass slab, aluminium sheet, grips – refrigerator (to store resins)

Microscopic Analysis

257

Fig. 8.1. (a) Preparation of thin section slides (equipment and materials required: resins, press, coating, UV polymerization apparatus, grinding stone cutting saw, grinding polishing machine) (b) Computer controlled station for automated modular preparation (Struers)

Materials and Reagents – Silicon carbide sheets (Carborundum–SiC) MOHS hardness: 9 – boron carbide sheets (B4C) MOHS hardness: 9 – corundum sheets (Alumina – Al2O3) MOHS hardness: 9 – diamond sheets (range of particle sizes) – polishing cloth – diamond water–ethanol soluble pastes and sprays: 3, 1, 0.25 µm – cold epoxy resins + hardener + catalyst + thinner – acrylic resins with low shrinkage rate – cold polyester resins – dyestuff for resin

258

Mineralogical Analysis

– canada balsam (+xylene) – lubricants and solvents: glycerol HOCH2–CH(OH)CH2OH, oil, ethanol, O–xylene C6H4(CH3)2, styrene C6H5CH=CH2, 1–4 dioxane, acetone, 99% diethylene triamine H2N–CH2CH2NH–CH2CH2NH2 – silicone grease for grinding The choice of a resin will depend on its: – refraction index (near 1.54) – solubility in an organic solvent or water – low level of contraction and low viscosity – polymerization conditions – hardness and strength at the required temperature particularly for observations and quantifications by electronic microscopy of the SEM and EDX type. Procedure Compact Hard Samples (Rocks) Petrographic techniques are described in detail by Hartshorne and Stuart (1970). The sample is suitably oriented and sawn on one face, then gradually polished and finally stuck on a petrographic slide of format 28 × 48 mm (1 in. × 2 in.). After sawing off a section 2–3 mm thick parallel to the stuck face, thin to 30 µm. Polish the thin section to 1 µm. The thin section slide can be protected from abrasion and oxidation by being stuck onto a thin glass sheet with Canada balsam. The slide should not be covered if subsequent tests require specific dyes, chemistry (elimination of carbonates) or samples are required for SEM-EDX measurements. Frangible, Porous or Fissured Samples Samples removed in the field with their natural moisture using cylinder or monolith methods of soil analysis (Pansu et al. 2001) are often oriented vertically in the profile or possibly oriented according to the field slope (Fig. 8.2), these should be stored in airtight boxes then used to produce the 28 × 48 mm standard or Mammoth slides. Whenever the block is being cut during preparation, the direction of the field micro-section should be taken into account (Fig. 8.2). The sample is dried by freeze-drying or by replacing the water with acetone or dioxane. Slight contraction will allow the sample to be released from the mould unless the system of sampling with two half-cylinders is used.

Microscopic Analysis

259

Fig. 8.2. Pedological micro-section

The block should be cut into a rough square with a cutter to reduce its thickness and, depending on the chosen orientation, deposited in a semiflexible plastic mould (in practice, the bottom of the mould corresponds to the bottom of the profile). Impregnation Place the block of soil with its mould in a vacuum desiccator with a ground lid lubricated with silicone grease. Put the desiccator under vacuum at approximately 60 cm Hg (or more) depending on the boiling point of the solvent and the resins. The soil should be degassed for 30 min. In a flexible plastic container, quickly mix the resin with the hardener (in the proportions recommended by the manufacturer) and the same volume of solvent to fluidify the mixture. Pour the resin mixture into the funnel at the top of the desiccator. Introduce the mixture gradually while monitoring vacuum, input flow and the rise in the level of the impregnation liquid until the sample is completely covered (plus one cm to allow for subsequent retraction of the resin). The mixture must also flow around the sample, and impregnation should take place upwards by capillarity. Depending on the porosity and size of the sample, 5–8 hours are often needed for 48 × 28 mm slides. Break the vacuum with care and transfer the sample under a fume hood. The thinner will take approximately one week to evaporate. The product becomes increasingly viscous, and then starts to harden (this will take about a month depending on the resin and dilution). Put it in a ventilated drying oven at 45°C for 2 or 3 days until the sample is no longer sticky. Unmould it respecting the orientation. Cutting Place the block on a bed of coloured resin (approximately 2 mm thick) that hardens rapidly in order to be able to identify the bottom of the soil

260

Mineralogical Analysis

profile. Saw the hardened impregnated block according to the preferential orientation (a slight notch can be cut in the side of the block to identify the direction of the field slope). The circular saw should be lubricated with a solvent that cannot dissolve soluble salts and should be compatible with the resin and with any chemical analyses to be performed later on. When impregnation appears to be complete, cut the block into parallel sections 5–6 mm thick. Each section should match the format of the petrographic slides. Rapidly check the quality of the surface for impregnation defects like cracks or flatness. The surface may need to be impregnated again after cleaning with alcohol and compressed air. Using a spatula, coat the section with a mixture of resin plus catalyst and place it on a flat glass block; cover it with a thin aluminium film pressing to exclude any air bubbles. After solidification, uncover the sample, trim it with a scalpel and leave it to harden in a ventilated drying oven at 45°C. The section is then ready for polishing. Using a silicon carbide abrasive disc with a rather coarse particle size, polish until almost complete abrasion of the re-impregnation layer, then, after cleaning with a blast of compressed air, continue polishing with a sequence of diamond abrasive paste of increasing smoothness (6, 3 and 1 µm). The surface must be perfectly smooth with no abrasions and no residues of the re-impregnation film. The material will look dull on the polished resin. Sticking the Sample on a 48 × 28 mm Petrographic Slide Clean the slide with solvent and dry. Write the slide number with India ink. Mix the resin used to fix the sample with hardener (at 60°C or cold depending on the nature of the sample and of the resin). With a spatula apply the resin to one surface of the holder slide and to the polished section of sample, and then rotate the two faces slowly one against the other moving them very gently to push out any excess adhesive and eliminate air bubbles. Allow to harden and trim the edges. After 48 hours, smooth the external face of the sample by sawing it to approximately 300 µm thickness, then by polishing it with set of diamond powders of decreasing particle size until reaching 30 µm (the usual thickness of thin slides for optical observation gives an orange birefringence in quartz). Each time the particle size of the diamond paste is changed, carefully clean the slide to eliminate all coarse particles which could scratch it. If minerals that are rich in iron are abundant, 30 µm is not thin enough to enable observation and it is necessary to reduce the section to approximately 20 µm to make it sufficiently transparent. For scanning electron microscopy or EDX probes, the final polishing should be done with a 0.25 µm diamond paste. For high resolution

Microscopic Analysis

261

transmission electron microscope observations, the glass support should be removed and the thickness of the micro-zones (diameter 3 mm) reduced, these should be separated and cut with a scalpel. An argon gun is used for thinning. For observation of minerals oriented in the same plane as the slide, a microtome can also be used for cutting. Microdiffraction will provide some information, but interpretation is often difficult on very thin samples. Separation on slides of micro-particles of about 50–100 µm is possible with an ultrasound probe equipped with a carbon needle on a micromanipulator; the slide is observed on a reversed optical microscope. Micro-particles are placed on a silicon plate and are analysed by XRD with slow scanning. The resulting spectra are compatible with angular spaces and standard intensities. 8.2.3 Grids and Replicas for Transmission Electron Microscopy Principle The samples can only be observed in transmission electron microscopy if they are less than 1 µm thick and if they are assembled on very thin conducting films. The object subjected to electron bombardment must be crossed by the electron beam. The acceptable thickness depends on the energy of this beam i.e. approximately 0.2 µm penetration at 50 kV, and 3–4 times higher at 100 kV. The beam–matter interaction heats the sample-target which can cause volatilization of the pore water; the morphology of halloysites is then seriously damaged and there will be contamination of the gun of the microscope. With elements of higher atomic number, heating can be very intense (fusion, phase shift, sublimation, destruction of supporting film). The supporting film should be transparent to the electrons, sufficiently solid to support the sample, resistant to heat and to electrostatic charges; it should be no thicker than about 100 Å. Reagents – Polyvinyl formal (FORMVAR, nD20: 1.50) – dichloroethane, Cl–CH2–CH2–Cl – 0.15% FORMVAR in dichloroethane solution; – COLLODION (nitrate of cellulose or pyroxylene C12H16N4O18) – amyl acetate (CH3–CO2–C5H11) or butyl acetate (CH3CO2(CH2)3CH3) – 1% collodion solution in amyl acetate – 1% fluorhydric acid in water.

262

Mineralogical Analysis

Equipment

Fig. 8.3. Grid of transmission electron microscopy φ 3 mm (322 meshes) numbers: location of x - coordinate letters: location of y - ordinate

– Lab glassware – Cu (or Ni) grids 3.05 mm in diameter (Fig. 8.3); these are also available covered with an FORMVAR–carbon film or with a perforated FORMVAR–carbon–gold film enabling direct observation of microparticles – 110 mm non-magnetic stainless grips with ultra-fine points – plastic film for replicas 0.034 mm thick – micro-drying oven to melt the replicas at 45°C – sampling loop. Procedure Grids – Place the required number of 3 mm grids on a glass slide previously moistened with water. Place the slide in a cupel and very gradually submerge it; the grids should not float – add a drop of FORMVAR–dichloroethane solution to the surface of the water; the drop should form a very thin film (<100 Å) without folds or holes which will solidify in the air after evaporation of the solvent (the film will be thinner if the water temperature is close to 0°C); eliminate the liquid slowly to bring the film into contact with the slide and the grids it covers; let dry then cover the grids with flash carbon which reinforces the film and makes it conducting; when the grids are grey, they are ready to use; check their quality under a binocular microscope

Microscopic Analysis

263

– dilute a drop of sample suspension in water to obtain an almost clear liquid; treat with ultrasound to separate the particles; remove one micro-drop and place it in the centre of the grid1 the grid should not be turned over during the operation; let dry at air temperature and dehydrate in a critical point apparatus if necessary. Replicas Samples which change shape during desiccation are too fragile or too dense to transmit the electrons but can be studied using a one or twostage replica technique that results in a slight decrease in resolution. First the sample should be rapidly subjected to directional shading with gold or platinum, then the surface covered with a vertically applied carbon film that is both conducting and resistant. In this case, cover the sample with a special 0.034 mm thick thermofusible film that softens at 45°C and preserves a replica of the surface of the mineral. Dissolve the clay sample with its cover in a diluted hydrofluoric acid solution. The replica remains in the hollows of the sample and can be subjected to gold or carbon treatment (if this has not already been done). Then dissolve the polystyrene film in ethylene chloride. Assemble the carbon replica on a grid and observe using TEM. All these procedures should be carried out with extreme care. 8.2.4 Mounting the Samples for Scanning Electron Microscopy

Principle The samples may be massive and rough. Depending on the characteristics of the sample vacuum chamber, discs 20 cm in diameter and 4 cm in thickness (8 in. wafers) can be used, but the degasification of large samples is only possible in the case of compact rocks with a limited number of fissures. In practice, it is preferable to use smaller samples. The surface for observation must be a clean break in the sample to enable the study e.g. the plan of cleavage, crystal orientations, defects in the crystal lattice, the presence of occluded impurities. Surfaces are rendered conducting with flash-carbon or by metallization if micro-probe analysis is not required. Otherwise, if the SEM is
1

The grids should be handled with forceps with ultra-fine points. 0.1 mL microsyringes of the Hamilton type. Flame-drawn hydrophobic glass tubes treated with PROSYL 28 can also be used.

264

Mineralogical Analysis

equipped with an EDX or WDX micro-probe, flat, perfectly polished surfaces (0.25 µm) are required. Equipment – Special SEM supports – storage boxes – 5 × 5 mm calibration grids with 2 µm squares – 3 mm carbon slides to mount on SEM supports – pencil marker for SEM (conducting ink) – double-face self-adhesive ribbon with low content of volatile elements. Products – Silver lacquer – conducting carbon lacquer – a set of reference minerals for quantification. Procedure

Fig. 8.4. Position of the samples on SEM supports

Small samples can be assembled on aluminium supports by sticking them together with silver or carbon lacquer. In certain cases, carbon supports can be used, or plates of 3 mm thickness stuck on aluminium supports. Carbon lacquer is generally preferable for EDX micro-probe analysis. It is essential to locate the samples precisely (e.g. by squaring, marking, or marking the right direction on the support); it is also important to avoid creating a vacuum under the sample (Fig. 8.4) because this causes discontinuity, and elimination of the charges can be disturbed resulting in scratches on the images which renders the photographs unusable. The lacquer should not cover the sample or fill the cracks. After prolonged drying to eliminate solvents, cover the samples with

Microscopic Analysis

265

flash carbon using a metal sprayer with ionic bombardment, or shadow with a metal deposit from an evaporator. 8.2.5 Surface Treatments (Shadowing, Flash-carbon, Metallization)

Vacuum Evaporator Flash carbon is often used to reinforce the FORMVAR film which supports the samples; it also makes the film conducting. Flash carbon should be applied vertically and uniformly. The micro-samples are sometimes only slightly absorbent and are not very visible in TEM. In this case the sample can be covered with a directional deposit of carbon whose grain is not very apparent, or be metallized with platinum or gold, under a tangential entry. Each space protected by a relief will appear shadowed. Knowing the angle of incidence, it is possible to measure the length of the shadow and deduce the height of the corresponding relief (Fig. 8.5).
Fig. 8.5. Vacuum evaporator (shadowing with flash carbon or metal)

Sputtering Metallization The apparatus consists of an anticathode made of gold shaped in a ring whose internal diameter is longer than the length of the sample support. At the centre, a cylindrical magnet is connected to a magnetic field which forms the other pole and surrounds the anticathode (Fig. 8.6). The electrons, which would otherwise overheat the sample, are deviated by the magnetic field.

266

Mineralogical Analysis

Fig. 8.6. Sputtering metallization apparatus (Bio-Rad – olaron) for SEM samples. N magnet, 1 Au anticathode, 2 sample support, 3 cooling block, filled circle neutral atoms, open circle positive ions. When the sample advances in the magnetic field, all sides are bombarded

The sample support is cooled to 4°C by a Peltier thermoelectric system. Heating is thus reduced and it is consequently possible to treat organic samples. The treatments are carried out under 10–1 Torr vacuum for 30–180 s. The applied voltage can reach 3 kV. Sweeping with a dry neutral gas (argon) enables elimination of residual traces of water, carbon dioxide, oxygen and possible oil contamination from the vacuum pumps. It is sometimes necessary to degas porous samples for several hours. In this way contamination is limited, but it is nevertheless often preferable to dehydrate on the apparatus at critical point before continuing degasification in the metal sprayer and then metallization. Cryo-fixing is often useful for organic matter and very frangible samples. This treatment gives excellent surface conductivity and accentuation of the relief of the rough samples by shadowing in SEM.

Microscopic Analysis

267

8.3. Microscope Studies

8.3.1 Optical Microscopy Description Optical microscopes allow observation of objects that are too small to be observed with the human eye or with a magnifying glass. Direct observations are carried out under IR to UV radiation including visible radiation, and can be accompanied by photography on film (black and white or colour) or digitalized video images. The magnifying power of a magnifying glass ranges from 2 to 60 times and of the most powerful microscopes up to 1,500 times. The object can be massive or very thin (a thin slide that is covered or not) to determine properties of soils or soil materials using absorption-transmission. Covering slides protects the surface from oxidation. Covered slides can be used for optical observation in immersion but the slides cannot subsequently be used for electronic microscopy. Briefly, an optical microscope is composed of a stand which supports a mechanical mount ensuring vertical displacement of an objective, and an eyepiece over an object slide. The magnifying power and the diameter of the field characterize the relations between the image and the object. Lightness refers to the luminosity of the optics used. The depth of field and the focusing range, as well as the limit of resolution determine the zones where the object can be observed under optimal conditions for a given material. A system of lighting allows observation by reflection or transmission. The lighting can be directed for massive objects (low-voltage lamps, optical fibres). For very thin objects, the lighting can be concentrated into a point by condensers with respect to different backgrounds: pale background, dark background, polarized light, phase contrasts, UV (slides covered or not). IR microscopes use optical systems with mirrors to avoid adsorption of IR by the materials generally used in the manufacture of lenses. Polarizing Microscope In soil sciences, polarizing microscopes are primary tools for the observation of crystals and the characterization of their optical properties. Interference microscopes or phase contrast microscopes (invisible transparent objects against a pale background) are rarely used. Variations

268

Mineralogical Analysis

in transmission factors can reveal structures that are invisible in natural light and make it possible to identify phenomena of pleochroism, isotropy and anisotropy of structure and mineral associations (as in forms, facies, cleavages, macles) using cross or slightly uncrossed polarizers. These microscopes (Fig. 8.7) enable observation of variations in transmission compared to the direction of polarization of the incidental light. A calibrated compensator placed in front of the analyser allows observations to be quantified. The choice of the objective is important (magnifying power, immersion or not) for the quality of soil observations.

Fig. 8.7. Diagram of a polarizing microscope

For the study of the structure and porosity of soils, the size, shape, associations of individual grains, and distribution of the different phases are determined either on extracted phases, or on thin sections (Jongerius et al. 1972; Bullock and Murphy, 1980). Procedure A rotating support, graduated in degrees, makes it possible to measure the extinction angle and to identify primary minerals that are still present as

Microscopic Analysis

269

well as to specify the scale and type of weathering. Interpretation requires considerable experience meaning only specialists in petrography are usually able to do it. Orthoscopic methods are time consuming but generally more precise than conoscopic methods (Wahlstom 1969, Hartshorne and Stuart 1970). This type of analysis is qualitative, but can be quantified by counting the mineral particles originating from the parent rock and minerals liable to weathering in the fractions previously separated by fractionation using particle size, density, hardness, magnetic properties, etc. The shape of the particles (rounded edges, sphericization of softer minerals), surface appearance (such as flatness of the particles, cleavages, cracks, different coverings) are indicators of the form of deterioration, erosion and transport (chemical weathering, micro-corrosion, waterice- or winderosion). The nature of minerals can be deduced from their colour, opacity, and especially from their refractive index and observable modifications in polarized light (e.g. pleochroism). Certain minerals have a more or less clear birefringence. The extinction angle can be reached with varying degrees of rapidity, and may be partial or complete. The shape of certain crystals is characteristic. These observations can be supplemented by scanning electronic microscopy after rapid mounting of the materials on double face sticking supports made conducting with flash-carbon (cf. Sect. 8.2.5). Using thin sections allows observation without disturbance of the in situ arrangement of the sample, orientations and associations of minerals, discontinuity of the mineralogical composition of a profile (decrease in or disappearance of certain minerals, ratio of minerals resistant to weathering:minerals liable to weathering giving index of deterioration). The development of concretions, nodules, the appearance of cementing, the presence of organic matter at different stages of decomposition can be observed and quantified by subsequent measurements using SEM combined with EDX (certain artifacts of preparation, like the filling of neocracks, or holes made by polishing by alumina are easily revealed). The units of organization (skeleton, plasma, vacuums) can be studied in detail at different scales using extracted fractions or/and thin sections: the skeletal components correspond to particles that are not reorganized, plasma corresponds to fine elements that can move and reorganize (such as clays and oxides), vacuums are related to porosity (circulation of air and solutions in soil). For the specific study of pore spaces, fluorescent colours can be mixed with the impregnation resin during the preparation of thin sections (cf. Sect. 8.2.2).

270

Mineralogical Analysis

8.3.2 Electron Microscopy, General Information All electron microscopes are based on the interaction of electrons with matter. The energy of an electron accelerated by a voltage V is equal to E = m v2/2 = e V (with m, v, e = mass, speed and charge of the electron, respectively). High energy radiation (fast electrons) can affect the level of the deep electronic layers of the atoms. Weak energy radiation (slow electrons) only affects the external electronic layers which reflect the chemical state of the atoms. The total effect of the electron beam is related to the electronic cloud of the Z electrons (e–) of the electronic orbitals around the nucleus of atoms. The following factors should be taken into account when considering how to change the way radiation affects matter: – intensity: transmitted or reflected intensity is lower than incidental intensity, absorption occurs – direction: there is scattering with loss of energy (inelastic scattering modifying internal structure), or without loss of energy (coherent elastic scattering allowing diffraction) – energy: as some energy is lost, reflected, transmitted or scattered energy is lower than initial energy. Losses in intensity and energy may be accompanied by modification of the matter due to the effect of the radiation: – in the case of electron microscopes with very high energy (3,000 kV), the sample can gradually be destroyed – in the case of microscopes with energy lower than 400 kV, there is transfer of energy by excitation of the electrons, thermal vibrations, particle ejection, and emission of secondary radiations usable for quantification. The heating effect produces phonons. Some chemical effects are reducing (e– gain). Chemical bonds can be ruptured. When the transfer of energy is higher than the threshold of displacement (between 15 and 30 eV), the effects of irradiation can cause atomic displacements. With electronic corpuscular incidental radiation of sufficiently high energy, an orbital electron in the deep atomic layers can be ejected with a kinetic energy corresponding to the difference in the energy lost by the incidental radiation and the electron’s own energy (secondary electrons). As the excited state is unstable, the atom subsequently returns to a fundamental state; there is release of X-photons or Auger electrons (relaxation phenomena). With electromagnetic radiation such as incidental or re-emitted X-rays photoelectrons are obtained (IR- to UV-photons, cathodo-luminescence). Electron microscopes can be classified at two levels depending on the geometry of the sample:

Microscopic Analysis

271

– massive samples which, as they are very thick, can be analysed only by the signals that come from their surface by reflection – samples with a critical thickness that allows radiation to cross them (micro-crystals, thin films, etc.), in which case measurements can be made by transmission. However, progress in instrumentation has led to changes in this dichotomy with the appearance of hybrid apparatuses allowing measurements on thin samples that use both processes. The following types of equipment are available: – traditional transmission electron microscopes TEM (possibly with additional functions in transmission scanning mode) – scanning transmission electron microscopes (STEM) – microscopes with scanning by reflection (conventional scanning electronic microscopes: SEM) – microscopes with scanning by reflection with differential vacuum where the observation chamber is under partial vacuum (environmental scanning electronic microscopes: ESEM). Each type of apparatus allows complementary measurements. The apparatuses are suitable for either very high resolution, or, with multiple configurations, for a range of different chemical and physico-chemical approaches. The signals obtained are complementary in the energy fields. 8.3.3 Transmission Electron Microscopy, Micro-diffraction

Principle Transmission electron microscopes use an incidental electron beam which, while crossing a very thin sample, provides information on the shape and structural distribution of elementary soil particles. The interaction of the electron beam with the matter results in images and micro-diffraction spectra (and enables selection of elementary chemical analyses as complements to the different electronic micro-probes in STEM).

272

Mineralogical Analysis

(Coil)

Limit of resolution B’ A’

Fig. 8.8. Comparison of an optical microscope and a transmission electron microscope

A”

B”

The geometry of an electron microscope can be compared to that of an optical microscope (Fig. 8.8). A source of electrons (high voltage electron gun) replaces the source of photons. A system of illumination with electromagnetic condensers concentrates the electron beam on the object; an electronic objective forms an intermediary image which is captured by projection lenses to form the final image on a fluorescent screen or a photographic device. Not all the radiations generated by the incidental electronic beam (Fig. 8.9) are used, since the apparatuses generally have only 1–2 sensors.2 X-photonic radiation (1, Fig. 8.9) can be collected by an EDX) detector (with Si–Li detection, or diodes without windows for analysis of light elements such as nitrogen and carbon), or WDX (with crystal monochromator).
2

At their maximum configuration, some top-of-the-range commercial analyzers include up to five sensors.

Microscopic Analysis

273

IR–UV–visible photonic radiation (2, Fig. 8.9) can be detected by cathodoluminescence. The scattered electrons (5) can be analysed by electron energy loss spectrometry (EELS) and enables analysis of light elements in STEM. Back-scattered electrons (6), secondary electrons (7), and Auger electrons are used for SEM images and transmitted electrons (12) for TEM and STEM.

Fig. 8.9. Types of radiations emitted during the bombardment of a sample by an electron beam

The choice of a transmission electron microscope depends on the nature of the observations required (magnifying power, high resolution, the need for high voltage for excitation or penetration, possible chemical quantification) and the cost, but also the versatility and the possibility to up-grade the equipment, its ability to cover the whole range of magnifying power including weak magnifying power, the degree of automation and ease of use, the quality and the cleanliness of the vacuum, contrast and performance against a dark background, possible coupling with systems for chemical analyses and image analysers, etc. The annual cost of maintenance contracts and consumables should also be taken into consideration, as these can represent 3–6% of the initial purchase price.

274

Mineralogical Analysis

Emission of electrons An electron gun is the source of electronic radiation. Generally, radiation is caused by the thermoelectronic emission of a filament of tungsten heated to 2,500°C (or of a tip of lanthanum hexaboride heated to 1,600°C) which forms the cathode (Fig. 8.10).
Fig. 8.10. Thermal-emission electron gun by: (1) tungsten (on the left or LaB6 (on the right) filament, (2) focusing electrode polarized negatively with respect to the filament (Wehnelt), (3) anode, (4) crossover (10 to 50 µm). Wehnelt is carried to a negative potential of a few volts to push the emitted electrons out of the axis and to concentrate them in a narrow beam

The tungsten filament has a diameter of approximately 0.1 mm, is V shaped and pointed at the end to focus the emission. The filament is heated to a high temperature under vacuum and is subjected to a work voltage of 4.5 eV. The conduction electrons can then cross the barrier of potential and an electronic cloud is formed. These electrons are accelerated by a potential V0. An electron beam is obtained whose energy is E0 = eV0. The emitted electrons move into the column at constant speed thanks to the high electric potential between the filament and the anode (supply voltage of the anode). A cathode made of lanthanum hexaboride (LaB6) with a work voltage of 2.7 eV, i.e. weaker than that of tungsten, can be used at a lower temperature (approximately 1,600°C); brightness is then considerably improved. However, the vacuum must be changed to 10–7 Torr, and the reactivity of LaB6 with certain metals can be awkward. Electron guns with field emission are also available whose brightness is much greater than that of thermoelectronic guns and whose energy dispersion is reduced. The incidental beam of electrons emitted by the electron gun (<1 mm of the cross-over) crosses the column of the microscope following the optical axis. Electromagnetic lenses are solenoid and consequently generate a magnetic field that focuses the electrons. An external shield prevents the dispersion of this magnetic field. The usual acceleration voltage varies from 50 to 1,250 kV, but can reach 3,000 kV. In practice, microscopes are available with (1) voltage of less than 100 kV, (2) medium voltage of between 200 and 500 kV, (3) high and very high voltage electronic microscopes (HVEM), 1,250 kV and above. Those in group (3) are very expensive, very voluminous and require special safety equipment.

Microscopic Analysis

275

Observations In TEM mode, high resolution electronic microscopes (HREM–HRTEM) (200–300 kV) can be equipped with 15 Å probes which enable the study of the morphology of the samples at different scales, direct observation of the atomic structure of a crystal and of the stacking of atoms (1.5 Å at 400 kV) on micro-samples where XRD is not efficient. These microscopes are thus useful to study problems of fundamental crystallography, phenomena of deterioration (germination and crystal growths, transformation of the phase that is amorphous to X-ray into crypto-crystalline and crystalline phases in the repetitive structures of clays). For example, interstratifications of mica–chlorite and minerals of 7–14 Å were studied by Amouric (1987, 1990) and Amouric et al. (1988), mica–kaolinite associations by Ahn and Peacor (1987).3 Using these techniques, it is possible to detect the planar defects, relic layers, and pale fringes of the interlayer levels. Care should be taken with high resolution to ensure that the high electron energies do not cause irradiation damage due to powerful vibrations of electron matter. Micro-diffraction In mineralogy, micro-diffraction of electrons is generally carried out simultaneously with the observation of images of normal incidence. The objects are prepared on micro-grids at sufficiently low density to insulate the elementary particles in the same way as for imagery. Micro-diffraction can be performed with the majority of the TEMs simply by adjusting the diaphragm. The fast electron beam at low wavelength and high energy (20–60 keV or more) strikes the extremely thin micro-crystal (<100 Å); when the angle of incidence on the reticular levels is in agreement with Bragg’s law (cf. Chap. 4) spots of diffraction are observed (Fig. 8.11). On submicro-samples, the spectra obtained are characteristic of single-crystal structures. Such a detailed view of crystal arrangements and defects cannot be obtained with traditional XRD (cf. Chap. 4).

3

For these very fine studies, the zones of interest are selected on thin slides with SEM at magnifying powers of about 10,000–20,000. These zones are separated on sections thinned with an ultra-microtome and an argon gun to ensure they are sufficiently transparent for the electrons in HRTEM.

276

Mineralogical Analysis

Fig. 8.11. Micro-diffraction by adjustment of the diaphragm

Micro-diffraction by adjustment of the diaphragm uses a diaphragm placed in the image plane which defines a reduced active surface of approximately 1 µm2. This method enables readable spectra of oriented micro-crystals to be obtained, but the very small wavelength of the electrons induces weak angles of diffraction and associated intensities that are different from the methods of traditional X-ray diffraction described in Chap. 4. As the sample is very thin, the diffraction spots obtained are characteristic of single-crystal structures. This method is often used as a complement to traditional XRD. On polycrystalline materials like clays, annular spectra can be obtained in about a minute. Special TEM Techniques Visualization of Charges with Colloidal Gold Principle As colloidal gold is opaque to electrons, it is used as a tracer to reveal edge charges or structural defects in crystalline structures (Photo 8.1 left). Equipment and reagents – TEM grids 3.05 mm in diameter covered with an FORMVAR film and carbon – 5 nm particle size colloidal gold solution (store in the refrigerator protected from light).

Microscopic Analysis

277

Photo 8.1. Special transmission electron microscopy techniques. On the left, highlighting of edge charges in kaolinite by colloidal gold (see procedure in text of this section), on the right view in paraglyph (see procedure in below), photographs (x 90,000), Gautheyrou J., IRD mineralogical reference set, Bondy, France, unpublished data

– Suspend the colloidal gold solution by agitation – take 0.5 mL of gold suspension and mix in a glass tube with 0.5 mL of sample of low density in order to obtain well separated minerals; leave in contact for a few minutes; agitate and put a micro-drop on a 3 mm grid., Allow to dry in the air and view under a TEM with a magnifying power of about 60,000. Gold preferentially migrates towards the rupture or crystallization zones and reveals the modes of assembly and the active sites of certain clays. Development in Paraglyph Principle The aim is to obtain a pseudo relief by superposition, with a tiny shift of negative and positive transparencies of the same image (Photo 8.1 right).

278

Mineralogical Analysis

Equipment and products – Negative film with strong contrast, format at least 6.5 × 9 cm – transparent positive film – photographic development products (developer – fixer). Procedure – Choose a clear negative of the image – by contact trace the image on a positive transparency of similar density – superimpose the negative and positive and find the optimal shift needed to obtain an effect of relief – draw by tracing or by enlargement. This type of image makes it possible to see coverings of particles more clearly and the effect of relief can be spectacular. Opacification of Samples that are “Transparent” to Electrons Minerals rich in iron are very opaque to electrons and can cause problems if they are too thick as the resulting images are very strongly contrasted and no details are visible. On the other hand, certain very fine minerals like allophanes are practically transparent to electrons if they are present in low concentrations. These preparations can be opacified with a lead salt (PbCl2 at 1% in water). The sample is left in contact with lead solution for one hour, then washed, suspended again and diluted to prepare a TEM grid. Scanning Transmission Electron Microscopy The electron gun and the condenser system used to produce the electron beam are based on a principle that is similar to traditional TEM, but in TEM the signal is transmitted to the image plane observable on a fluorescent screen via a system of electronic lenses, whereas in scanning transmission electron microscopes (STEM), the signal is directly collected by electron or X-ray detectors, and transmitted on-screen (Fig. 8.12). In true STEM, an electron gun with field emission, whose cross-over is about 5 nm and whose brightness is more than 1,000 times higher than that of a traditional tungsten source, provides an electron beam which crosses a condenser giving a reduced image of the source (micro-probe).

Microscopic Analysis

279

Fig. 8.12. Diagram of a scanning transmission electron microscope

This probe scans the surface of the sample by means of a deflecting coil. The electrons transmitted or diffracted by the sample are collected on a detector with a response that is proportional to their intensity. After amplification, an image is created on the screen stage-by-stage in synchronization with the scanning generator. Electron guns with field emission are very sensitive to contamination. They require an ultrahigh “dry” vacuum (10–10 Torr), which proscribes the use of oil diffusion pumps for the secondary vacuum. In spite of the use of cryoscopic traps, the gun can still break down because of traces of oil. Very high spatial resolution can be achieved. This equipment can also be equipped with energy analysers such as electron energy loss spectrometers (EELS). They can carry out analyses on surfaces of the order of one nanometer on all the elements of the periodic table (3Li to 92U) on submicroscopic samples. Dedicated STEM are not the most widely used, many manufacturers prefer to sell hybrid TEM equipped with complementary STEM which perform excellently for a price that is 2 at 3 times lower.

8.3.4 Scanning Electron Microscopy Scanning Microscopes by Reflection, Microprobes The concept of the scanning electron microscope (SEM) and that of electronic micro-probes (EM) are complementary, EM comprising probes of less than 1 µm optimized for X-ray analysis.

280

Mineralogical Analysis

The thermionic electron gun is subjected to a negative voltage of 10– 50 kV. The sample is placed on a goniometric precision support (a binocular magnifying glass enables visual location of the point of impact on the microprobe). Electromagnetic condensers form the image of the probe which is projected on the sample. The probe is moved by deflection of the beam. A massive sample can be 2–4 cm thick and have a diameter of 20 cm (8 inches wafer). A large-capacity sample chamber requires a clean vacuum system with a strong flow (turbomolecular pump). The magnifying power is the ratio of the amplitude of the scanning of the image (fixed) to the amplitude of the scanning of the object (variable). Electron–matter interactions (secondary and back-scattered electrons, X-ray, Auger electrons, photoluminescence, transmitted electrons) can be used for analytical measurements. The image is created stage-by-stage (pixel by pixel) and allows digitalization and treatments using an associated data processing system. The creation of the images is based on two modes: – in secondary electrons mode the incidental primary radiation of the electrons loses energy in contact with the matter; part of the energy is restored in the form of secondary electrons which cross the grid of the collector and are then accelerated in the field of the scintillator; an exploitable signal is obtained which is mixed with the back-scattered electrons which are able to cross the diaphragm of the detector – in back-scattered electrons mode the electrons are collected by the collector of a scintillation detector; the signal is rather weak but detection is improved by using a semiconductor detector in the shape of disc that is perforated in the centre which is placed above the sample; a device installed in two or four different sectors makes it possible to create a topographic contrast. The chemical composition of the sample sometimes varies in a random way because the rate of penetration is very low. The shade of grey is related to the atomic numbers of the elements observed. The resolution is about 20–100 Å depending on the element observed. The intensity of the electron beam and of the scanning conditions is chosen to ensure maximum resolution and an optimal signal-to-noise ratio for a given magnifying power. Even in the best conditions strong incidental energy (approximately 30 keV) prevents very fine details from being observed, but generally reasonably good results are obtained. On the other hand, if the material is slightly conducting and cannot be sprayed with metal, it may be better to reduce the charge by using energy below 5 keV.

Microscopic Analysis

281

The diaphragm should be selected to obtain a suitable depth of field, as well as to allow adjustment of the work distance if the relief of the sample is significant. Environmental Scanning Electron Microscopy These microscopes enable high resolution images to be obtained by reflection on samples preserved in their natural moisture, without degassing or surface conducting treatment. Some environmental investigations can be made without deformation or transformation of the sample. Two systems are used: – low vacuum scanning electron microscopes (LV-SEM) are relatively simple and can be used in conventional SEM; they enable a partial vacuum of about 2–4 Torr to be created in the sample chamber; they are generally equipped with a detector of back-scattered electrons – environmental scanning electron microscopes (ESEM) are dedicated microscopes which enable a high vacuum to be created in the electron gun (10–7 Torr) and simultaneously a reduced vacuum of near atmospheric pressure to be created in the observation chamber. This difference in vacuum is obtained in stages with progressive reduction in pressure. The distance between the sample and the output of the electron beam under high vacuum must be as small as possible in order to avoid a reduction in performance. In conventional SEM, more than 95% of the electrons do not undergo dispersion. In environmental SEM with the sample at a short distance from the beam output under a pressure of 1 Torr, the proportion of non-dispersed electrons can reach 90%, but decreases with the number of gas molecules in the trajectory of the beam. A specific gaseous secondary electron detector (GSED) enables the quality of the image to be improved by discriminating the back- scattered electrons and the secondary electrons resulting from the interactions of the electrons of the beam and the atoms of the sample. There is no artefact of charge as in conventional SEM (e.g. ionization of gas, production of free electrons, or creation of positive ions compensating for the negative charges). This detector is not sensitive to light or to temperature. The atmosphere in the sample chamber can be controlled at the same time as the pressure and the temperature and enables observations in a gaseous medium of almost constant composition. Interpretation of the images requires adaptation to phenomena such as condensation on the minerals (for example rounding of the angles), the presence of interstitial

282

Mineralogical Analysis

water or pollutants and the determination of gas balances. Quantitative measurements by EDX (cf. Sect. 8.3.5) are possible. Many applications in soil science, especially in studies of organic matter, clayey materials, and micro-organisms are now possible using ESEM (Mathieu 1998, Leroux and Morin 1999), for example: – physical problems involved with expansible minerals, allophane soils with high water retention, structure, texture, porosity, aggregates, transfers between the soil and the environment, dehydration and hydration processes, soil shrinkage, compression, adhesiveness – effect of heat or chemical treatments, fusion, sublimation, growth of crystals, stabilization of structure, tests under constraint – dynamics of the degradation of organic matter, micro-fauna . 8.3.5 Ultimate Micro-analysis by X-Ray Spectrometry

Energy Dispersive X-Ray Spectrometry Micro-determinations are usually carried out by X-ray fluorescence spectrometry (cf. Sect. 31.3.2, Chap. 31) by means of an EDX spectrometer. This system (Fig. 8.13) enables plotting of charts of elementary qualitative distribution at the surface layer and on approximately 1 µm thickness. It is better to use almost flat surfaces; for accurate quantitative analysis, surfaces have to be polished to 0.25 µm to limit possible topographical effects. A fixed probe can be used if less precise quantitative micro-analyses is needed than that obtained with dedicated analytical probes, but ZAF matrix-correction software (Z: atomic number, A: absorption, F: fluorescence) enable improvement of the results. These analyses can only be performed on elements heavier than 11Na. Light elements require the emission of Auger electrons; but the ultra-high vacuum of 10–10 Torr required in Auger spectroscopy cannot be obtained with normal scanning microscopes. A SAM4 microscope is required where the vacuum is obtained with an ionic pump. Wavelength Dispersive X-Ray Spectrometry The source of X-rays emitted at the electron beam–matter interface is placed on a focusing circle called “Rowland circle” (Fig. 8.13). Detection
4

SAM = Scanning Auger Microscope.

Microscopic Analysis

283

is carried out by moving the crystal analyser and the entry slit of the detector (counter with proportional action) along the circle. The detector must be at the effective focal spot (2θ compared to the incidental beam).

Fig. 8.13. Microprobes with dispersion of energy and wavelength: EDS, EDX: energy dispersive X-ray spectrometry, WDS, WDX: wavelength dispersive X-ray spectrometry (Rowland circle, effect of the defocusing of the electron beam, θ = Bragg angle, ∆θ = deviation of the Bragg angle caused by defocusing)

To carry out quantitative analysis, the direction of measurement and the opening must be constant. In practice, the angle of reflection cannot exceed the 5–70° range. It may thus be necessary to use four crystal analysers (Microspec-USA system) at different reticular distances (Bragg’s law) to cover the range of wavelengths accessible with this approach (lithium fluoride, LiF, Pentaerythritol, PET, rubidium acid phthalate, RAP, lead stearate, STE). The WDX system is tending to be replaced by the faster EDX system.

References
Ahn JH and Peacor DR (1987) Kaolinization of biotite : TEM data and implications for an alteration mechanism. Am. Miner., 72, 353 Amouric M (1987) Growth and deformation defects in phylllosilicates as seen by HRTEM. Acta Cryst., R43, 57 Amouric M (1990) Etude de l’interstratification mica-chlorite par microscopie électronique. In Matériaux argileux, structure, propriétés et applications,

284

Mineralogical Analysis

Decarreau A ed., Soc. Fr. de Minéralogie et Cristallographie, Groupe Français des Argiles, 283–287 Amouric M, Bianetto T and Proust D (1988) 7.1 and 14 Å mixed layer phyllosilicates structurally studied by TEM in pelitic rocks. Bull. Miner., 111, 29 Bullock P and Murphy CP (1980) Towards the quantification of soil structure. Microscopy, 120, 317–328 Delvigne JE (1998) Atlas of micromorphology of mineral alteration and weathering, The Canadian Mineralogist, special publication 3, Ottawa, IRD, Paris Hartshorne NM and Stuart A (1970) Crystals and the polarizing microscope., Arnold, 219–251 Jongerius A, Schoonderbeeek D, Jager A and Kowalinski S (1972) Electrooptical soil porosity by means of Quantimet B equipment. Geoderma, 7, 177–198 Leroux A and Morin P (1999) Evolution de la microscopie à balayage – un progrès pour les applications géo-environnementales. Bull. Lab. Ponts et Chaussées, 222, 85–89 Mathieu C (1998) Effects of electron-beam/gas interaction on X-ray microanalysis in the variable pressure SEM, Microchim. Acta., 15, 295–300 Moran C., McBratney AB and Koppi AJ (1989) A rapid analysis method for soil macropore structure. I. Specimen preparation and digital binary image production. Soil Sci. Soc. Am. J., 53, 921–928 Pansu M, Gautheyrou J and Loyer JY (2001) – Soil analysis – sampling, instrumentation and quality control, Balkema Publishers, Lisse, Abington, Exton, Tokyo, 512 p. Tessier D (1985) Validité des techniques de déshydratation pour l’étude de la micro-organisation des sols. In Soil Micromorphology, Fedoroff N., Bresson LM and Courty MA ed., AFES Wahlstrom EE (1969) Optical crystallography., Wiley New York

Chronobibliography
Deer WA, Howie RA and Zussman J (1962, 1963, 1966) Rock-forming minerals., vols. 1–6, Longmans-Green London Beutelspacher H and van den Marel HW (1968) Atlas of electron microscopy of clay minerals and their admixtures., Elsevier Amsterdam Reid WP (1969) Mineral staining tests. Mineral Ind. Bull., 12, 1–20 Spry A (1969) Metamorphic textures., Pergamon, Oxford Hartshorne NH and Stuart A (1970) Crystals and the polarising microscope., Arnold Gard JA (1971) Electron-optical investigation of clays., Mineral Society Hutchison CS (1974) Laboratory handbook of petrographic techniques., Wiley New York Wells OC (1974) Scanning electron microscopy., McGraw-Hill New York Brewer R (1976) Fabric and mineral analysis of soils., Krieger USA

Microscopic Analysis

285

Goldstein JI and Yakowitz H (1976) Practical scanning electron microscopy., Plenum New York Zussman JB (1977) Physical methods in determinative mineralogy., Academic New York Tessier D and Berrier J (1979) Utilisation de la microscopie électronique à balayage dans l’étude des sols. Observation des sols humides soumis à différents pF. Sci. du Sol., 1, 67–82 Jogerius A and Bisdum EBA (1981) Porosity measurements using the quantimet 720 on backscattered electron, scanning images of thin sections of soils. In Submicroscopy oth soils and weathered rocks., Centre Agr. Pub. and Doc., Wageningen, Pays-Bas Smart T and Tovey NK (1981) Electron microscopy of soils and sediments, t1 et t2. Clarendon Oxford Fleischer M, Wilcox RE and Matzko JJ (1984) Microscopic determination of opaque minerals. US Geol. Survey Bull., 1627 Low AJ, Low EJ and Douglas LA (1984) A motorized grinder for making soil thin sections. Geoderma, 32, 335–337 Bullock P, Fedoroff N, Jongerius A, Stoops G, Tursina T (1985) Handbook for Soil Thin Section Description, Waine Research. Publication. Wolverhampton, England Goldstein JI, Newbury DE, Echlin P, Joy DC, Fiori C and Lifshin E (1985) Scanning electron microscopy and X-ray microanalysis., Plenum New York Maurice F, Keny L and Tixier R (1985) Microanalyse et microscopie électronique à balayage., Les éditions de Physique Murphy CP (1985) Fasten methods of liquid-phase acetone replacement of water from soils and sediments prior to resin impregnation. Geoderma, 35, 39–45 Willaime C (1987) Initiation à la microscopie électronique par transmission en minéralogie, science des matériaux., Soc. Fr. de Minéralogie et de Cristallographie Blackburn M, Caillier M, Bourbeau GA and Richard G (1988) Utilisation d’une solution de chlorure de sodium pour le remplacement de l’eau dans les échantillons d’argile lourde avant l’imprégnation. Geoderma, 41, 369– 373 Takeda H (1988) A rapid method for preparing thin sections of soil organic layers. Geoderma, 42, 159–164 Chartres CJ, Ringrose-Noase AJ and Raupach M (1989) A comparison between acetone and dioxane and explanation of their role in water replacement in indisturbed soil samples. J. Soil Sci., 40, 849–863 Wright D, Stanley D, Chen HC, Shultz AW and Fang JM (1990) A frame based expert system to identify minerals in thin section. Microcomputer applications in geology II. Pergamon Oxford, 289–299 Zhurov AV (1990) Preparation of polished sections for the study of soil pores and their differentiation by size. Pochvovedenize, 8, 144–147

286

Mineralogical Analysis

Sludzian G and Galle P (1992) Cartographie de matériaux et d’échantillons biologiques par microscopie ionique à balayage. La vie des sciences, Compte rendu série générale, 9, 157–177 Gribble CD and Hall AJ (1993) Optical mineralogy : principles and practice., Chapman and Hall London Fitzpatrick EA (1993) Soil microscopy and micromorphology., Wiley Ringrose-Voase AJ and Humphreys GS (1994) Soil micromorphology, studies in management and genesis ; Proceedings of the IX international working on soil micromorphology., Elevier Science Ltd Lavoie DM, Little BJ, Ray RI, Bennett RH, Lambert MW, Asper V and Baerwald (1994) Environmental scanning electron microscopy of marine aggregates. J. Microscopy, 178, 101–106 Vempati RL, Hess TR and Cocke DL (1996) X-ray photoelectron spectroscopy. In Methods of soil analysis, Sparks DL ed., SSAA book series No 5 Jonhson RA (1996) Environmental Scanning electron microscopy – An introduction of ESEM, Philips Electron Opt.,-FEI, 55p Mathieu C (1996) Principle and application of the variable pressure SEM, Microscopy and analysis, 43, 13 Pichler H, Schmitt-Riegraf C and Hoke L (1997) Rock-forming minerals in thin section., Chapman, London Hitachi, (2000) Low temperature microscopy using a cooling stage. Hitachi technical data, 62 Astley OM, Chanliaud E, Donald AM and Gidley MJ (2001) Structure of Acetobacter cellulose composites in the hydrated state. Int. J. Biol. Macromole, 29, 193–202 Slowko W (2001) Secondary electron detector with a micro-porous plate for environmental SEM, Vacuum, 63, 457–461 Tai SSW and Tang XM (2001) Manipulating biological samples for environmental scanning electron microscopy observation. Scanning, 23, 267–272

Part 2

Organic A nalysis

9

Physical Fractionation of Organic Matter

9.1 Principle and Limitations
9.1.1 Forms of Organic Matter in Soil Many different organic fragments can be distinguished in soils, most of which are of plant origin (living or dead roots, fragments of wood fibre, fragments of stems and dead leaves) and others of animal origin (e.g. cadavers, faecal pellets, earthworm casts) from macro-fauna such as insects, arachnida, myriapodes, crustacea, gasteropods or earthworms. Increasing the scale of observation, smaller organic fragments e.g. filamentous roots, partially decomposed animal debris, nematodes, fungi, and algae can be identified with a magnifying glass. The observation of other micro-organisms (e.g. bacteria, actinomyces, protozoa) and debris of animal or plant origin that are increasingly incorporated in organomineral colloids requires a higher power of magnification. Initially soil organic matter (SOM) thus appears to be a continuum of increasingly fine fragments that can be physical fractionated. 9.1.2 Principle The methods of fractionation described in this chapter include both manual or mechanical sorting, and the use of physical techniques for density fractionation, particle-size fractionation by sieving, and analysis of sedimentation. The methods used resemble those used in fractionation of mineral particles (cf. Chap. 2).

290

Organic Analysis

Manual sorting is used especially for studies on live roots in the soils. Sorting is facilitated by floating using – for example – water elutriators (cf. Sect. 9.2.2). Density fractionation is based on the difference in density between matter of plant origin (close to 1) and of mineral origin (around 2.65 for primary minerals). Theoretically, it is thus an ideal technique for the separation of fragments of plant origin that are not decomposed in the soil. This type of measurement is now widely used in studies on the dynamics of carbon in soils. Some compartment models were established with data obtained by densimetric separation (Pansu and Sidi 1987, Arrouays 1994). These techniques are described in Sect. 9.2.4. However, depending on the type of soil, the density fractionation method may be hindered by the close associations between mineral and organic particles. The technique can be improved by combining it with particle size fractionation (Sallih and Pansu, 1993) and with a range of dispersion techniques described in Sect. 9.2.3 of this chapter. The aim of particle-size fractionation is complete separation of the organic components of the soil. Ideally, coarse fractions of more than 50 µm would contain intact plant debris, silt fractions of 50–2 µm (or 20–2 µm) would contain cells and microbial fragments, coarse clays of 2–0.2 µm would contain organic matter of the organomineral complex, and finally fine clays of 0–0.2 µm would contain recently formed metabolites. Some approaches, such as dating (Anderson and Paul, 1984) or isotopic 14C measurements (Hassink and Dalenberg 1996) or δ13C (Puget et al. 1995) appeared to partially confirm this theory. Other studies showed that micro-organisms and organic materials are closely associated with mineral colloids, and consequently “clean” fractionation of the biological components of soils is impossible (Ahmed and Oades, 1984). Certain review studies (e.g. Christensen 1992, Feller 1994) did, however, identify three main classes of organic matter: – a plant debris compartment (>20 µm) that is not closely associated with mineral sands with a relatively high C:N ratio (15–25) or with a high xylose to mannose ratio, indicating that this organic matter is of plant origin – an organic silt complex including a mixture of soil organic matter (SOM) of plant and fungal origin, mineral silts and very stable organomineral micro-aggregates; C:N and xylose:mannose ratios are lower than in the previous compartment; the origin of this organic matter is not as clear as that of coarser matter

Physical Fractionation of Organic Matter

291

– an organic-clay compartment (<2 µm) rich in amorphous SOM; this compartment is humified and closely associated with the mineral particles; C:N (8–11) and xylose:mannose ratios are lower, suggesting that the origin of this organic matter is probably microbial. 9.1.3 Difficulties Particle-size fractionation uses sieving techniques (generally wet sieving) to separate particles until 50 or 20 µm. The separation of the finest particles requires sedimentation techniques similar to those described in Chap. 2 for mineral fractionation. However, an additional difficulty to take into consideration is particle density with respect to the Stokes law of sedimentation (cf. Chap. 2). The average density of 2.65 used for mineral particles is not appropriate for organic fragments (Elliott and Cambardella 1991). However, at this particle-size, organic matter is often closely associated with mineral particles. As SOM content is relatively low compared to mineral particle content, one can consider that the densities of the mineral fractions are not significantly modified. The main difficulty in physical fractionation by density or particle size separation lies in the close association between minerals and organic matter resulting in different types of aggregates (Fig. 9.1). These aggregates have to be broken down and the organic components released without destroying them. Sect. 9.2.2 of this chapter discusses various techniques of dispersion at some length, and describes their limits and comparative interest. Preparation and especially rewetting of the sample involves a risk of modifying the organic constituents. The samples are generally dried before storage at the laboratory. Drying, together with other preparation techniques (Pansu et al. 2001), stops the organic functioning of the soil resulting in deactivation or death of micro-organisms. The treatment is brutal and it is preferable to slow down microbial activity by cold storage or even better by freezing the fresh soil. Whatever the technique of conservation used, stopping the biological activity is necessary to preserve the soil organic state at sampling, as the kinetics of evolution of certain organic components are higher than that of the main inorganic components. However, rewetting the soil starts biological activity again. Rapid growth of the populations of micro-organisms supplied with the plant debris which are released from their clay protection during the preparation of the samples (effect of grinding) and become available for micro-organisms, as well as by consumption of the microbial biomass

292

Organic Analysis

killed during these operations. Part of this carbonaceous source is then mineralized or transformed. There is thus a risk of changing the organic contents of the soil by rewetting of the samples which is required by most of the physical fractionation techniques described later. However, this risk is limited in the presence of a great excess of water, since most active food chains are essentially aerobic. The risk can be also limited by the use of reagents that are unfavourable to the growth of micro-organisms and by performing fractionation as rapidly as possible after moistening.

Fig. 9.1. Formation of organomineral complexes, micro-aggregates and structural aggregates (after Bruckert 1994)

Physical Fractionation of Organic Matter

293

9.2 Methods
9.2.1 Classification Methods for physical fractionation of organic components can be classified in three main groups: – separation of the plant roots; – separation by density; – fractionation in particle-size ranges. When using these methods, the structure of the soil material should be kept in mind i.e. humified fine organic matter associated with inorganic matter forming organomineral complexes. These complexes involve bonds between solid particles resulting in the formation of different types of aggregates in the soil. Some of the components that have to be separated are imprisoned in these aggregates (Fig. 9.1). The difficulty in fractionation thus consists in splitting up these aggregates without destroying the components that have to be measured. The principal techniques for aggregate dispersion are described and commented in Sect. 9.2.3. 9.2.2 Extraction of Plant Roots Objective and Principle This type of extraction is useful to measure root production in the soil. Indeed the production and turnover of roots is one of the most significant inputs of carbon in the soil, the other inputs being root exudation and above-ground necromass production by the plant. The study of the carbon balance in the soil and in the atmosphere has been the object of intensive research and methodological compilations e.g. Anderson and Ingram (1989) that describe methods to estimate organic inputs in the soil. Root extractions are also useful for observations of plant physiology such as classification of roots, estimation of their weight and length, chemical analyses, biological associations with fungi and bacteria. Separation is generally carried out manually but several types of elutriation apparatuses are available. Procedures Extraction is performed on intact samples from blocks or cylinders of soil. The samples must be stored in polyethylene bags at low temperatures or even better frozen. If a freezer is not available they can be dried and rewetted before washing, but the best approach consists in

294

Organic Analysis

washing the roots immediately after returning to the laboratory from the field. In addition to organic matter content, the texture of the soil, compaction, and structure affect the difficulty of the extraction to a varying extent. The simplest method consists of gently washing the wetted samples with water on a sieve whose mesh size differs with the author: 2 mm for Abo (1984), 0.5 mm for Anderson and Ingram (1989); The material remaining on the sieve can be washed with water and separated by decantation. To remove all the fragments, the residue often has to be sorted manually under water in flat containers. This work may require a binocular magnifying glass and very fine forceps. The difficulty of the work also depends on the type of soil and roots. Many machines have been described that wash roots; most separate roots from soil by elutriation, i.e. washing the debris accompanied by their separation by flotation on a 0.5 mm sieve located far from the heavy particles. Fig. 9.2 shows a diagram of the apparatus designed by Smucker et al. (1982) which is based on the principle of hydropneumatic elutriation. The apparatus built by Bonzon and Picard (1969) is suitable for the separation of roots from intact soil sampled in the form of cylinders or monoliths. It is composed of a set of 4 double sieves made of wood with a brass screen with a rectangular section (50 × 60 cm) and a cylindrical bottom. The top sieve has a mesh of 1.18 mm, and the bottom sieve of 1.4 mm. The sieves are set in a wooden frame that is moved backwards and forwards at 12.5 oscillations min−1 by an engine with a crank-connecting rod system. Samples of a volume of around 2 L are placed on the top sieve and jet water is directed onto the sample. Slurry is evacuated over the top of the raised edge of the top sieve. Once washing is complete, the contents of the sieves are transferred to a funnel equipped with a sieve with a very fine mesh. This funnel contains the organic fragments but also stones and gravels with a diameter of over 1.4 mm. If there is a lot of gravel, the organic fragments should be separated using a strong jet of water directed at the base of the funnel and transferred onto a sieve placed below. After mechanical fractionation of soil and roots, it may be necessary to manually sort the roots from the other organic debris and this operation can take several hours. Consequently there is no “ideal” machine that eliminates all manual operations. All separation methods result in losses of fine roots and washing water and residues should be checked periodically to quantify these losses.

Physical Fractionation of Organic Matter

295

Fig. 9.2. Diagram of an apparatus for the separation of the plant roots from soil by hydropneumatic elutriation (after Smucker et al. 1982). A: high energy water washing chamber, B: elutriation chamber, C: transfer tube, D: first sieve with weak kinetic energy (840 µm), E: second sieve (420 µm). Tube C is separated from B for cleaning and to introduce a new sample. The roots are transferred from the weak energy sieve (D) by reversal and washing onto the fine sieve (E)

Soaking for one night in an aqueous sodium hexametaphosphate solution accelerates the process of washing roots in clay soils, but the chemical action may discolour the roots and break down certain plant tissues thereby rendering subsequent identification of live roots impossible. This type of pretreatment can also interfere with chemical analysis of the roots. In addition, all these treatments can damage the contents of tissues and it is consequently preferable to separate a subsample of roots by hand and to wash the roots carefully with a minimum of water to enable accurate chemical analysis. The washed roots can be stored in the refrigerator in sealed polyethylene bags but freezing is preferable. A small quantity of bactericide such as thymol can also be added.

296

Organic Analysis

Dry matter weight and organic carbon and nitrogen content (cf. Chap. 10) can be measured after drying at 70°C for 48 h. Bonzon and Picard (1969) also measured the specific surface of the roots in addition to dry weight. Progressive calcination up to 550°C with successive temperature steps enables determination of the ash weight of the roots and quantification of the inorganic elements after dissolution of the residues in acid solution. 9.2.3 Dispersion of the Particles Structure of the Soil and Organic Components As mentioned in the introductory section to this chapter, soil always contains varying proportions of coarse materials of inorganic (coarse sands, gravels) or organic origin (plant fragments), in addition to structural aggregates whose form and stability vary with the type of soil. In active medium (mull), humification processes result in relatively large quantities of transformed organic matter: microbial metabolites with a rapid turnover (e.g. many polysaccharides), and very stable phenolic products, both of which are accounted for by SOM decomposition models (Pansu et al. 2004). Both types of materials can bond to mineral matter to form organomineral complexes such as the cements contained in soil micro-aggregates. These micro-aggregates also comprise the building materials of larger aggregates containing organic particles, organic debris and microbial species (Fig. 9.1). In soils with a low level of activity (moder, mor), the formation of a strongly differentiated profile with a resistant organic matter horizon (moder, mor) is likely, along with a horizon in which redistributed organic matter accumulates resulting in the organomineral complexes of the structures of precipitation (Bruckert, 1994). Fractionation thus depends on the different forces of cohesion of the soil structure. In certain cases, simple moistening is enough to break down the macro-aggregates and disperse the fine particles (slaking). In other cases, more energetic dispersion techniques are needed to release the micro-aggregates and the organic fragments embedded in the structural aggregates.

Physical Fractionation of Organic Matter

297

Dispersion Techniques Dispersion consists in breaking certain organomineral binding forces without fragmenting plant debris, and if possible avoiding damaging microbial cells or the structure of the micro-aggregates. It should be kept in mind that the aim of granulometric fractionation of organic matters is very different from particle-size analysis of soil (cf. Chap. 2). In particle-size analysis, very energetic methods are used to destroy clayhumic “cements” e.g. destroying organic matter with hydrogen peroxide, destroying organomineral bonds with reagents that are highly complexing for iron and aluminium such as sodium tetraborate or sodium pyrophosphate. Such techniques are not appropriate here since the organic components need to be recovered without them being damaged or dissolved. The most useful methods can be classified in three groups: – dispersion with water and possibly with mechanical agitation of varying strength – sonic and ultrasonic dispersion – chemical dispersion with dispersing reagents that are not too aggressive for organic matter. Mechanical Dispersion with Water Bruckert (1994) recommended this type of dispersion technique rather than techniques using ultrasounds whose action varies considerably with the type of organomineral cement of the aggregates and is considered to be too destructive for some soil compounds as ultrasounds break fragile minerals and damage certain organic matter, but especially “cause the breakdown of the microbial cells from which the protoplasmic contents come to be adsorbed on clays (Mc Gill et al. 1975)”. The technique of Bruckert et al. (1978) is a low-impact mechanical treatment by controlled agitation in the presence of agate balls (35 g of dry soil sample from the fine earth prepared at 2 mm, 200 mL water is placed in a rotary shaker with five agate balls and agitated at 50 rpm for 15 h). Feller (1979) developed a similar technique on tropical sandy soils with low humus content. In this case the recommended mechanical action is even more moderate: 100 g soil agitated for one hour with three glass balls in 300 mL distilled water. Andreux et al. (1980) studied a standard steppe soil of the chernozem type with a very stable clay–humus complex. The dry soil was sieved to 2 mm, shaken by slow rotation (40 rpm) in water (35 g of soil for 200 mL water) for one night at 20°C with different numbers of agate balls. The rates of the fine clay–silt fraction (0–50 µm) obtained by these authors increased from 57% of the soil weight with agitation without agate balls

298

Organic Analysis

to 90% of the soil weight when mechanical fragmentation was used. Beyond two, the number of balls had a limited influence on the rate of the fine fraction (Fig. 9.3). On the other hand, up to 15 h of agitation the rate increased without reaching the next stage, revealing that destruction of all aggregates bigger than 50 µm is progressive. However, beyond a certain degree of mechanical action (more than three balls for 15 h of agitation or five balls for more than 8 h of agitation), the treatments appear to solubilize part of the carbon, so very aggressive mechanical action is not recommended.

Fig. 9.3. Influence of the number of 10 mm agate balls (a), and length of agitation (b) on the fragmentation of the aggregates >50 µm (after Andreux et al. 1980) filled diamond mass percent of the 0-50 µm fraction, filled square carbon content of the 0–50 µm fraction as a percentage of total C

Sidi (1987) also used fragmentation by agitation with glass balls on a Tunisian carbonated soil. Fig. 9.4a shows the influence of the length of agitation on particle size distribution with or without the presence of three glass balls (15 g of soil, 100 mL water, back and forth agitation with one back and forth movement per second. The main effect of the mechanical treatment was the destruction of the biggest macro-aggregates (200–2000 µm) whereas the percentage of aggregates of intermediate size (50–200 µm) remained almost identical with or without the balls. The shape of the curves also suggested that the process of fragmentation occurs in stages (1) division of the biggest aggregates (>200 µm) into intermediate aggregates (50–200 µm) during the first 30 min of agitation, followed by (2) division of the 50–200 µm aggregates into microaggregates of the size of clays and silts (0–50 µm). In contrast to the situation illustrated in Fig. 9.3, one hour of agitation with three glass balls was enough to reach a dispersion plateau for this type of soil. Monnier et al. (1962) recommended performing dispersion procedures before densimetric fractionations (cf. Sect. 9.2.4): either by dry

Physical Fractionation of Organic Matter

299

sieving to 500 µm, or boiling in water followed by one rinse in alcohol and one period in the drying oven.

Fig. 9.4. Dispersion of a Mediterranean soil with water (according to Sidi 1987, 15 g soil/150 mL water, back and forth shaking apparatus, 1 backwardsforwards movement per second) a: influence of length of agitation with and without three glass balls filled square, open square 200–2,000 µm with and without balls, respectively, filled triangle, open triangle 50–200 µm with and without balls, filled circle, open circle 0–50 µm with and without balls; b: influence of ultrasounds (80 W–80 kHz) on agitation for 1 h without balls

Sonic and Ultrasonic Dispersion Although occasionally severely criticized for being too destructive for certain organic matter (Bruckert 1994), sonic and ultrasonic dispersion techniques are generally recommended for the physical fractionation of soil organic matter. Edwards and Bremner (1967) subjected an aqueous suspension of the soil sample (10 g soil for 25 mL water) to sonic vibration (9 kHz, 50 W) with a Raytheon S-102A vibrator (Raytheon Co., Norwood, MA USA). For 14 soils of very different texture, dispersion in fine particles of the size of clays (<2 µm s) by sonic vibration for 30 min was evaluated by the pipette particle-size fractionation method (cf. Chap. 9). Fig. 9.5 shows that dispersion was always much higher than by simple agitation in water. Dispersion was comparable with that obtained with the two chemical dispersants tested: calgon peroxide and sodium resin. The rate of dispersion obtained on the suspensions with an ultrasonic probe MSE Cabinet Model 60 Ultrasonic disintegrator (Measuring and Scientific Equipment Ltd, London) delivering a frequency of 18–20

300

Organic Analysis

kHz and a power of 60 W, is very similar to that obtained by sonic vibration (Fig. 9.6). Beyond 30 min (the period recommended by the authors) the duration of sonification had only a slight influence on the percentage of clay obtained (Fig. 9.7), Fig. 9.4 b shows a comparison of the influence of the period of sonification observed by Sidi (1987) with a slightly more powerful high frequency ultrasound probe (80 kHz, 80 W); in this case a plateau was reached earlier (at 8–10 min) for the dispersion of the particles the size of silt (0–50 µm).

Fig. 9.5. Rates of clayey fractions obtained by ultrasonification and three other dispersion techniques on 14 soils (after Edwards and Bremner 1967)

After a detailed analysis by Watson (1971) of the ultrasonic vibration method applied to the dispersion of soils, Genrich and Bremner (1972) re-evaluated the technique following some criticism of its use. They used 28 soils covering a very varied field of pH (3.6–8.2), carbonate content (0–34% CaCO3 , texture (2–59% of sand, 7–72% of clay) and organic content (0.14–9.4% organic C). They tested two types of instruments (Heat Systems Ultrasonics Inc, Plainview, NY USA) (i) a standard Branson W-185C model with probe (20 kHz, 80W) and (ii) a Branson 220 ultrasonic cleaner with stainless steel tank (40 kHz, 100W). Different procedures were used with the tank model (soil:water ratios of the suspensions, sonification in an Erlenmeyer flask or directly in the

Physical Fractionation of Organic Matter

301

tank). With the probe model, the end of the probe (diameter 1.27 cm) was immersed to 2 cm below the surface of the suspension (10 g of soil in 25 mL water) in tubes of steel cooled on the outside to less than 20°C.

Fig. 9.6. Comparison of clay rates obtained on five soils by sonic (9 kHz, 50 W) and ultrasonic (18–20 kHz, 60 W ) dispersion for 30 min on suspensions of 10 g of sample in 25 mL water (after Edwards and Bremner 1967).

In all cases, more complete dispersion was obtained with the probe model than with the tank model. However, the dispersion provided by the probe depended to a great extent on the quality of its surface: with a pitted probe, the authors observed that the length of time needed for dispersion was two to four times longer than with a probe in good condition. It is consequently important to gently polish the end of the probe with a fine abrasive paper after each 30-min period of use. According to Genrich and Bremner, the imperfect condition of the probes could explain the failures noted by other authors before their trials. They also showed that with a 15-min period of sonification with the probe, under the conditions described above, the clay rates obtained on their 28 soils were always equal to or higher than those obtained with the sodium peroxide and polyphosphate method of Kilmer and Alexander (1949) which at that time, was the standard method of dispersion (Soil Survey, 1960). This study clearly demonstrated the dispersion power of ultrasonic probes. However, the authors’ conclusion was cautious saying that no method of dispersion can be described as universally applicable for all soils. Anderson et al. (1981) studied the distribution of organic matter in the particle fractions of two soils of the chernozem type. They carried out the dispersion of these soils by ultrasonic vibrations for 8 min with more power than previously (300 W, apparatus Bransonic 1510) but applied to more diluted suspensions (soil:water ratio of 1:10). Tiessen and Stewart (1983) studied the effect of cultivation on the organic composition of the particle fractions using a procedure similar to that of Anderson et al. (1981).

302

Organic Analysis

On a soil of the chernozem type, Shaymukhametov et al. (1984), like Anderson et al. (1981), observed a stage of fragmentation of microaggregates (<50 µm ) after sonification for 30 min, whereas in one minute, 96.4% of the larger aggregates were destroyed. Their experiment highlighted the very great difference in stability between microaggregates and structural aggregates (Fig. 9.1). It should also be noted (Fig. 9.7) that the degree of stability of the three sizes of microaggregates between 1 and 50 µm , is very similar, the probable explanation being that ultrasounds cause the progressive release of fine clayey particles from the three classes of the silt-size micro-aggregates with no distinction between the classes. This is different from the behaviour of structural aggregates where there is a clear difference in stability between the 50–200 µm and 0–50 µm fractions (Fig. 9.4).

Fig. 9.7. Effect of the duration of ultrasound treatment on fragmentation of microaggregates <50 µm (after Shaymukhametov et al. 1984): filled square <1 µm, filled diamond 1–5µm, filled triangle 5–10 µm, times 10–50 µm

In order to study the organic matter of an aquoll, Catroux and Schnitzer (1987) performed ultrasonic dispersion on soil in water suspensions at a ratio of 1:5 (between the ratios used by Genrich and Bremner, 1972, and by Anderson et al. 1981). 100 g of soil in 500 mL distilled water were agitated on a magnetic stirrer and treated by ultrasound with a Blackstone SS2 generator. Power was applied at 400 W for 15 min, (which is a more energetic treatment than that applied by the preceding authors) and the end of the probe was immersed to 2–3 cm below the surface of the liquid in order to decrease the swirling action.

Physical Fractionation of Organic Matter

303

Gregorich et al. (1988) tried to define and quantify the action of ultrasounds more precisely. Ultrasonic vibrations cause cavitation due to the formation of microscopic bubbles resulting from local reductions in pressure and the subsequent bursting of these bubbles. When the bubbles burst in the suspension, they produced waves of pressure of sufficient mechanical energy to break the aggregation bonds. These authors used a 20 kHz Branson probe whose power could be adjusted from 0 to 150 W. The probe head (diameter 12 mm) was immersed to between 5 and 10 mm below the surface of the suspensions (15 g of 1–2 mm aggregates in 75 mL water). The output power of the probe was gauged by measuring the rise in temperature of a known water mass over a given period. Gregorich et al. considered that the most significant parameter is the quantity of energy applied per mL of suspension: J = P t V –1 where J is the applied energy in J mL–1, P is the output power of the probe in W, t is the time in s, V is the volume of suspension in mL. Figure 9.8 shows the results obtained by these authors on a melanic humus gley horizon of a cultivated brunisol. This type of material has very resistant silt particles. None of the ultrasonic treatments enabled their fractionation as thoroughly as treatment by agitation in the presence of hydrogen peroxide. The principal bond between these particles thus appears to be primarily organic. These authors also observed stronger bonds between macro-aggregates (or pseudo-sands) than in the majority of studies quoted above, energy ranging between 300 and 500 J mL–1 being required to disperse these aggregates which are relatively rich in organic matter. As is true for silt particles, organic matter thus seems to act as cement, particularly in macro-aggregates. One possible explanation is that their organic functioning is a little different (partly anaerobic) in this type of soil from the examples above. Balesdent et al. (1991) studied the effect of ultrasounds on the granulometric distribution of the organic matter contained in 17 soils (Ap horizons of cultivated soils), type brown soils, or not very processed alluvial soils. The procedure they used combined mechanical and ultrasonic dispersion techniques. The first mechanical dispersion used rotary shaking of the aqueous suspensions with glass balls similar to the techniques described above (Andreux et al. 1980). Sonification was then applied to the fraction below 50 or 25 µm in order to split it into three

304

Organic Analysis

particle sizes: 0–0.2, 0.2–2, 2–50 µm (or 2–25 µm). The ultrasound apparatus used was the same type as the one used above (Branson cell disintegrator, 20 kHz, 150 kW, probe with a flat head 13 mm in diameter).

Fig. 9.8. Ultrasonic dispersion of a melanic humus gley horizon of a cultivated brunisol (after Gregorich et al. 1988): filled triangle >50 µm fraction, filled square 2–50 µm fraction, filled diamond <2 µm fraction; horizontal lines represent dispersions obtained after H2O2 treatment of destruction of organic matter

A kinetic study of the action of ultrasounds on silt-size microaggregates was performed by Balesdent et al. (1991). Sonification was applied to suspensions of 100 mL with a soil:water ratio of 1:3, the probe was immersed to 3 cm below the surface and the apparatus regulated at 70% of its power (corresponding to 0.5 W mL–1 from the manufacturer). Figure 9.9 shows changes in 0–0.2 and 0.2–2 µm fractions and their sum (0–2 µm) compared to the reference method (hydrogen peroxide treatment and pyrophosphate dispersion). Compared to the results of Gregorich et al. (Fig. 9.8), the dispersion of soils studied by Balesdent et al. appears to be easier as it has a stable production of 0–2 µm clay fraction from 300 to 1,800 J mL − 1 (Fig. 9.9). In comparison, the clay fraction in Fig. 9.8 there is a continuous increase with an increase in the energy applied. However, Fig. 9.8 shows a clean break in the slope of the surve of the clay fraction for an energy of approximately 300 J mL − 1, i.e. about the energy needed to reach the stage shown in Fig. 9.9. In a clay latosol, Roscoe et al. (2000) found that

Physical Fractionation of Organic Matter

305

energies of 260–275 J mL−1 were sufficient to break down unstable aggregates (2,000–100 µm) and to leave stable aggregates (100–2 µm) unchanged.

Fig. 9.9. Effect of the energy of the ultrasonic treatment on fragmentation of micro-aggregates the size of clays in an alluvial soil (1) and a weathered brown soil (2); horizontal lines represent dispersions obtained by chemical treatment with H2O2 then Na3PO4 (Balesdent et al. 1991): filled square 0–0.2 µm fraction, open circle 0.2–2 µm fraction, closed circle 0–2 µm fraction

It is difficult to compare the data of Balesdent et al. (Fig. 9.9) with that of Shaymukhametov et al. (Fig. 9.7) where the cutting threshold of the fine fractions was at 1 µm . However, in both cases, the length of sonification needed to reach the 0–2 µm and 0–1 µm stages was quite similar. The more detailed study by Balesdent et al. of the fractionation of the fine 0–2 µm fraction provided interesting additional information on two aspects: – even with the highest energy of sonification, a stable stage is not reached for the fine fraction below 0.2 µm , and the dispersion of this fraction is always lower than that obtained with the chemical method of reference; – on one of the soils, the intermediate fraction (0.2–2 µm) reached maximum after around 5 min of sonification (150 J mL−1). This suggests an initial stage in the fragmentation represented by the division of the aggregates of silt size (2–25 µm) into smaller units (0.2– 2 µm) rather than the fragmentation of the 0.2–2 µm fraction. The behaviour of associations within the clay-size fraction is apparently different from that observed within the silt-size fraction where the three particle-size ranges studied (Fig. 9.7) displayed the same stability. Instead it resembles that observed for macro-aggregates (Fig. 9.4): the

306

Organic Analysis

large structural aggregates (>200 µm) are less stable than the intermediate macro-aggregates (50–200 µm). Finally, for the soils they studied, Balesdent et al. recommended a sonification period of 10 min (600 J mL−1) in the conditions described above. Dispersion of the silt micro-aggregates (to 2 µm) can then be considered complete, whereas the coarse clay fraction (0.2–2 µm) must be considered as micro-aggregated. Balesdent et al. also studied the effect of ultrasounds on the coarse organic debris separated in water after the action of glass balls. The study was on an alluvial soil containing 27% clay, and 0.9% organic carbon with a pH of 7. Corn and maize had been grown on the soil for 17 years so the coarse fragments mainly came from these plants. The ultrasound treatment was applied at different energies to aqueous suspensions at a ratio of 1:200 of each of the three light fractions: 200–2,000 µm, 50–200 µm and 25–50 µm. The suspensions were then sieved at 25 µm and if necessary at 50 and 200 µm. The 0–25 µm suspension was separated by sedimentation into fractions of 0–5 µm and 5–25 µm. The results showed a very destructive effect of ultrasounds on the organic debris. After 10 min (the recommended time for fractionation of clayey particles), more than 60% of the carbon of the initial coarse organic fraction was split into the lower particle-size ranges, and this was the case for each of the particle-size ranges studied. These authors showed that part of this fractionation results from the separation of clay fractions associated with plant fragments; but cleaning of the plant fragments is insufficient to explain the quantities of organic matter transferred to the finer fractions. The use of ultrasounds under the conditions required for dispersion of clays produces marked fractionation of the coarse plant fragments. This significant observation led the authors to propose a procedure for particle size fractionation that uses only ultrasounds for the suspension of particles of less than 50 µm (cf. Sect. 9.2.4). Chemical Dispersion Chemical dispersion techniques are less widely used for organic fractionation than for classical soil particle size analysis (cf. Chap. 2). As mentioned in Sect. 9.2.2, sequestering reagents such as sodium tetraborate or hexametaphosphate can only be used to disperse clay soils when the aim is to recover roots or coarse plant fragments (Anderson and Ingram, 1989). But even in this case, there is a risk of

Physical Fractionation of Organic Matter

307

discolouration that subsequently makes it difficult to identify living roots, and of modification of the organic contents. Dispersing reagents that are highly destructive for organomineral bonds are not recommended for the study of particle-size distribution of organic matter. Their too high extracting power, in particular of humic and fulvic acids (cf. Chap. 11), is likely to distort the results of such studies. Less aggressive extracting reagents should be used such as monovalent neutral salts which cause the dispersion of clays by exchange with the di- or trivalent cations of the exchange complex and the consecutive rupture of certain organomineral links. Ladd et al. (1977), Oudinot (1985), Sallih and Pansu (1993) used a sodium bicarbonate solution as a complement to the mechanical action of agitation for the initial dispersion of the soils. Sodic resins have also been used for the dispersion of soils (Edwards and Bremner 1967, Rouiller et al. 1972). Adapted from studies by Edwards and Bremner, Fig. 5 shows that using the resin technique, dispersion is slightly higher than using the two other methods tested for most of the 14 soils in this experiment. Feller et al. (1991) compared different dispersion techniques, including an IRN77 amberlite resin in a sodic state. The resin was tested alone (R) or combined with ultrasonic fractionation on the fraction below 50 µm (R/US). The resin technique was compared with five other dispersion methods which were all combined with the same ultrasonic fractionation of the fractions below 50 µm: a B/US method similar to that of Balesdent et al. (1991) described below, an NaCl/US method replacing the water in the suspensions by M sodium chloride solution, an M sodium hydroxide method bringing the suspension to pH10 (pH10/ US) and a 3.3 g L 1 sodium hexametaphosphate method (HMP/US). Figure 9.10 shows the comparative effectiveness of the different methods on a ferrallitic soil from Martinique. Up to the level of fine silts, the most effective dispersion was obtained using the R/US technique (resin on the total soil then ultrasounds on the fraction below 50 µm). On average, solubilization of organic matter was less than 4% of total carbon in the 19 soils studied. Based on the results of this experiment, the R/US technique appears to be preferable to the technique using glass balls plus ultrasounds (B/US) described above (Balesdent et al. 1991). However, the two methods were not tested on the same types of soils. In addition, the authors mentioned practical constraints in the use of the resins: the time needed for resin regeneration and preparation is rather long and there is a risk of contamination of the soil by very fine resin (<50 µm).

308

Organic Analysis

Fig. 9.10. Effect of different dispersion methods on particle size fractionation of a ferrallitic soil from Martinique (according to Feller et al. 1991): US: ultrasonification of the 0–50 µm fraction, B: stirring with balls, NaCl: NaCl solution, pH10: NaOH solution, HMP: sodium hexametaphosphate solution, R: stirring with sodic resin

The sodic resin technique was also shown to be the most effective of the five dispersion techniques for stable oxisols with high gibbsite content (Bartoli et al. 1991). These authors also studied the influence of the soil:sodic resin ratio on dispersion, pH of the suspensions, and carbon solubilization. Volumes of 0, 10, 50, 100, 200, 300, 400 mL Amberlite IR-120 (500 µm) sodic resin in nylon bags with a 50 µm mesh were added to samples of 2.5 g soil in 200 mL distilled water. The suspensions were agitated for 16 h on a rotary shaker at 40 rotations per minute. The results in Fig. 9.11 show a stable stage of aggregate breakdown for volumes of resin ranging between 50 and 200 mL, this corresponds to the volume (100 mL) used by Feller et al. (1991). There was a rise of between one and two units in the pH of the suspensions; in all cases the final pH remained lower than that of the main extracting reagents of the humic acids (cf. Chap. 11). In the deeper horizon, dissolution of organic carbon only became perceptible with volumes of resin above 200 mL; on the other hand, in the cultivated surface horizon, the authors noted dissolution of organic carbon at all doses of resin independently of the

Physical Fractionation of Organic Matter

309

added volumes; this horizon probably contains more recently formed organic matter which is not very humified, and is water soluble.

Fig. 9.11. Influence of the volume of sodic resin (2.5 g for 200 mL distilled water) on dispersion of the aggregates, the pH of the soil suspension, and solubilization of organic carbon in a surface horizon (on the left) and a deep horizon (on the right ) of a Nigerian oxisol (Bartoli et al. 1991)

9.2.4 Separation by Density The Techniques The first methods used to separate the organic fragments in the soil were usually based on an obvious physical property: the density of free organic matter, which is close to 1, is lower than that of the organomineral complex. However, the first density techniques were used for the separation of primary minerals (Pearson and Truog, 1937). Starting from the work of Lein (1940), Hénin and Turc (1949) adapted a densimetric separation technique for free organic matter in soils. Fractionation was performed in beakers containing a mixture of bromoforme and benzene.

310

Organic Analysis

The technique was improved by Jeanson-Luusinang (1960) by the use of special decantation funnels, and then further improved by Monnier et al. (1962) who adapted an earlier technique of Lein (1940) for density separation by centrifugation. The centrifugation technique improved the use of differences in density, but neither of these two techniques extracts all the light organic matter. Monnier et al. (1962) carried out tests with synthetic mixtures of mineral soil and oat straw. For fragmentation of straw in particles of less than 0.2 mm, the recovery rate was 73% of the added straw with the technique of Monnier et al. and 44% of the added straw with the technique of Jeanson–Luusinang. The method of Monnier et al. was used to model the evolution of carbon stocks by Arrouays (1994). Greenland and Ford (1964) used ultrasounds to disperse the aggregates before density separations (cf. sect. 9.2.3). The technique was improved by Ford et al. (1969) with the use of surfactant and of dibromochloropropane (DBCP density = 2.06) instead of bromoforme for density separations. At that time authors were not concerned with the possible toxicity of these products, which today is widely acknowledged. Turchenek and Oades (1979) studied methods of density fractionation of organic matter by combining them with preliminary particle-size fractionations. They carried out from 4 to 7 density fractionations with mixtures using decalin (decahydro naphthalene d = 0.88), dibromochloropropane (DBCP d = 2.06) and bromoforme (d = 2.88) on most of the seven standard particle ranges (coarse sands, fine sands, coarse silts, fine silts, coarse clays, medium clays, fine clays). Their observations showed that more than 50% of the light fraction (d < 2.06) with a particle size of coarse and fine sands is made up of organic matter. The fraction comprising coarser particles is mainly made up of recognizable plant fragments with high C:N ratios and low solubility. The fraction made up of finer particles (fine sands to coarse clays) contains a higher proportion of identifiable microbial cellular debris and soluble aromatic humic compounds. The light clay fractions are also rich in organic materials. Forms of oxidized iron, aluminium and silicon are present to a significant degree in all the fractions, indicating a wide range of different interactions between inorganic and organic matter. Nowadays none of the density methods using chlorinated heavy solvents are used because of the toxicity of this type of solvent and of the safety requirements in laboratories. Dabin (1976) proposed a method for the fractionation of organic materials (cf. Chap. 11). The first part of this method comprised density

Physical Fractionation of Organic Matter

311

fractionation on phosphoric acid 2 M (d = 1.2). In addition to its low toxicity compared to density liquors, this type of acid treatment has the advantage of destroying carbonates in calcareous soils releasing a certain proportion of sequestered plant material. Sidi (1987) used this technique to separate light fractions from mixtures of soils and wheat straw incubated in controlled laboratory conditions. The method was used to propose a descriptive model of carbon dynamics with three compartments (Pansu and Sidi 1987). The Ladd et al. (1977) method includes a series of particle-size and density fractionations that are also suitable for calcareous soils. A modification of this method allowed, in its first stage, fractionation of the light materials from in vitro incubation experiments of mixtures of soils and 14C labelled wheat straw (Cortez 1989, Sallih and Pansu 1993). The suggested modification concerned the use of an aqueous saturated zinc sulphate solution (d = 1.4) as heavy liquid, whereas Ladd et al. had used carbon tetrachloride (d = 1.59). Among the different high-density saturated saline solutions possible, zinc sulphate and ferrous sulphate (density = 1.6) appeared to be particularly promising. A zinc sulphate solution was selected to avoid the sequestering of iron on the organomineral complexes. However, it is probable that the zinc element also results in the formation of certain complexes. Other mineral density liquors have also been used including zinc chloride solutions (Besnard et al. 1996), sodium metatungstate (Elliott et al. 1991), sodium polytungstate (Cambardella and Elliott 1993, Golchin et al. 1994, Six et al. 1999), Ludox, aqueous suspension of silica colloidal particles (Meijboom et al. 1995). Anderson and Ingram (1989) recommended methods of fractionation of light materials with water that are rather similar to those described for the extraction of roots. The light fraction is defined as the fraction (1) which floats when it is dispersed in water, (2) which passes through a sieve of 2 mm but not through a sieve of 0.25 mm. However these authors pointed out that elutriation and sieving methods separate significantly less free organic matter than density methods. The methods of separation in water are nevertheless worthwhile because less organic matter is solubilized in water than in dense liquors, which are often rather corrosive (Beare et al. 1994, Puget et al. 1996).

312

Organic Analysis

Procedures Only procedures for the methods of Monnier et al. (1962), Dabin (1976) and density liquor ZnSO4 are described here (cf. Sect. 2.4.1 “The Techniques” for modifications). Method Using Organic Heavy Liquid The density liquid should be adjusted to the density selected by mixing bromoforme with a lighter solvent, preferably alcohol (Monnier et al. 1962). The density recommended by the authors who worked on silt soils in the area of Versailles (France) is 2. These authors pointed out that more complete separation could be achieved by the successive use liquids of density 1.75, 2 and 2.25. The soil sample should be air dried and crushed to 2 mm particle size. Weigh 5–10 g of soil depending on the free organic matter content. The weight of the sample can be also adjusted as a function of the techniques to be used for quantification after fractionation (e.g. carbon determination on the whole light fraction). Place the sample in the 100 mL tube of a centrifuge, and fill the tube with the density liquid. After stirring with a glass rod, centrifuge for 5 min with an acceleration of about 1,000g in the centre of the tube. Collect the supernatant on a flat filter and repeat the operation again by suspending the centrifugation pellet in the heavy liquid. It is possible to destroy aggregates to release embedded light organic matters before fractionation either by boiling in water then washing with alcohol and drying in the drying oven, or by sieving the dry sample at 500 µm. In the case of soils with high free organic matter content, the risk of sequestration of dense particles within light organic materials is high and it is thus recommended to centrifuge the light materials again after washing in a different tube with the heavy liquid. Density Method with Phosphoric Acid The method of Dabin (1976) applies to soil sieved to 0.5 mm. The weight of the sample can vary from 5 g to 40 g depending on the organic content. Agitate for 30 min on a back and forth shaker (1 backwards–forwards movement per second) with 200 mL of a 2 M phosphoric acid aqueous solution (136 mL L−1). Centrifuge for 20 min at 3,000g then transfer the supernatant on a filter. Repeat this operation twice. Dry and weigh the plant matter recovered on the filter. The total carbon of this material can be measured by combustion and determination of released carbon dioxide; nitrogen can be measured simultaneously with a CHN analyser, or separately using the Kjeldhal method (cf. Chap. 10).

Physical Fractionation of Organic Matter

313

Method by Sieving and Inorganic Heavy Liquid Figure 9.12 summarizes the procedure for extraction of free organic matter (FOM). Sieve a 80 g soil sample on a 5 mm sieve and put in suspension in 300 mL of 0.2 mol (NaHCO3) L−1 aqueous solution. After 1 h of moderate agitation on a rotary shaker, centrifuge at 12,000g for 30 min. Collect the light fraction by filtering the supernatant. The extraction can be repeated twice.

-1

Shaking centrifugation filtration

sieving

d Shaking centrifugation filtration organic matter Continuation Fig 9.13
Fig. 9.12. Diagram of the separation of free organic matter by density and sieving

Suspend the centrifugation pellet in 500 mL water. For dispersion, place on a rotary stirrer for 2 min at maximum speed. Sieve on a 50 µm sieve in water. Suspend the coarse fraction (greater than 50 µm) in a heavy aqueous solution saturated in zinc sulphate (density 1.4). After 30 min of agitation at mean velocity, centrifuge at 3,000g, then filtrate the supernatant on a 3 µm Millipore membrane. Wash the light fraction recovered on this filter carefully four times with water, add to the previous light fraction; dry in the drying oven at 30°C.

314

Organic Analysis

If the sands do not have to be separated from FOM, the method can be simplified by leaving out density separation on the coarse fraction. In this case FOM is estimated by carbon determination on the fraction: “light matter separated on NaHCO 3 + fraction greater than 50 µm”. Certain types of soils studies with 14C tracers have shown that this simplified estimate is significantly more exhaustive than the preceding one (Sallih and Bottner, Cefe-CNRS Montpellier, France, unpublished data). 9.2.5 Particle Size Fractionations Limits of Density Methods The use of the density method on its own for the study of organic components has sometimes been criticized, but not when density fractionation was coupled with particle-size fractionation. One of the reasons mentioned above is the toxicity of heavy organic liquids. This obstacle can be overcome by using heavy aqueous solutions saturated with mineral salts. According to Bruckert (1994), using density as the only criterion can also be challenged for several reasons: – the ideal density to use varies with the type of soils. Thus, with a density of 1.8, 90% of the organic matter of andosols can be separated in the light fraction whereas in brown soils the percentage is only 20% – the density of the plant debris increases during decomposition by incorporation of mineral matter which can be determined by ash quantification – in the case of organic heavy liquids, organic compounds can fix on clays and perturb subsequent studies. As mentioned earlier, inorganic heavy liquids do not have this disadvantage, but they can modify organomineral complexes. Procedures for Particle-Size Fractionation Given the remarks quoted in Sect. 9.2.3 about aggregate dispersion, it is difficult to describe a single procedure for particle size fractionation for all soil types. However four procedures appear to be appropriate for different soil types: – the continuation of Section Methods by Sieving and Inorganic Heavy Liquid” adapted from Ladd et al. (1977) on calcareous soils – Agitation with Glass Balls and Ultrasonification (Balesdent et al. 1991) used on different cultivated soils of France ”

Physical Fractionation of Organic Matter

315

– “Resin H + and Ultrasounds” (Feller et al. 1991) used on tropical soils of various origins, with a simplified alternative for sedimentation (Gavinelli et al. 1995) – a special procedure for use on sandy soils (Feller 1979; Feller et al. 1991). Continuation of the Procedure Described in “Method by Sieving and Inorganic Heavy Liquid” (cf. Sect. 9.2.4) In addition to separation of the “free organic matter” fraction described in “Method by Sieving and Inorganic Heavy Liquid” (Fig. 9.12), this method provides: – a water-soluble organomineral fraction – a fraction of more than 50 µm (primarily inorganic, density >1.4) – an organomineral fraction with particles of less than 50 µm.
9.

centrifugation at 800g

-

Fig. 9.13. Fractionation by centrifugation of the clay–silt fraction (complement of Fig. 9.12, after Ladd et al.1977)

The complete method includes the separation of this last fraction into particles the size of silts (2–50 µm), coarse clays (0.2–2 µm) and fine clays (0–0.2 µm). The separation procedure described by Ladd et al. (1977) shown in schematic form in Fig. 13 should be used: (1) centrifugation for 15 min at 4,000g in a 250 mL tube makes it possible to separate the fine fraction (less than 0.2 µm) in the supernatant; (2) the centrifugation pellet is then suspended again with water and centrifuged

316

Organic Analysis

for 5 min at a low speed (800g); repeated twice, this operation makes it possible to isolate a centrifugation pellet of silt size (50–2 µm) and (3) a supernatant of coarse clay size (0.2–2 µm). The two clay fractions and the water-soluble fraction are concentrated in a vacuum rotary evaporator at 40°C. The method (Figs. 9.12 and 9.13) thus provides six fractions: a water-soluble 0–0.2 µm fraction, a 0.2– 2 µm fraction, a 2–50 µm fraction, a heavy coarse fraction (size > 50 µm and density > 1.4), and a light coarse fraction (size > 50 µm and density < 1.4). Dry each fraction in a Petri dish at a low temperature, depending on subsequent measurements either at room temperature (light matter) or in a drying oven or sand bath. Agitation with Glass Balls and Ultrasonification Figure 9.14 is synoptic diagram of fractionation after Balesdent et al. (1991). Dry the soils in air and sieve to 2 mm using a grinding–sieving machine with rollers (Pansu et al. 2001). Put a 50 g sample in a 250 mL plastic bottle with 180 mL of water and ten glass balls 5 mm in diameter. Agitate the bottle on a rotary shaker at 40 rpm for 16 h. Filter the suspension underwater on a sieve with a 200–µm square mesh. Put the nib in suspension in a beaker. The organic fragments are separated during their transfer to a 200–µm sieve by decantation. Repeat this operation several times until the sands no longer contain any visible organic fragments. Perform the same operation on the fraction of less than 200 µm with a 50–µm sieve to obtain F200–2000, M200–2000, F50–200, M50–200 fractions (F being the organic fragments, M the organomineral part). Centrifuge the suspension with particles <50 µm to separate the particles of less than 0.2 µm (cf. Continuation of the Procedure Described in); reserve the supernatant. Suspend the centrifugation pellet at a solid:water weight ratio of approximately 1:3. Subject the suspension to ultrasound treatment for 10 min under the conditions described in above i.e. at an applied energy of approximately 300 J mL–1. In samples containing limestone, it is recommended to eliminate carbonates in the suspension after ultrasound treatment by adding HCl solution to a pH of 3.5 on the pH-meter, and to wash the solid residue before subsequent separations. The 2–50 µm, 0.2–2 µm and 0–0.2 µm fractions are separated by centrifugation techniques similar to those described earlier. The conditions chosen here are only slightly different from those of the previous authors: 25 min at 2,900g to separate the fine fraction <0.2 µm by decantation and 3 min at 800g for the 0.2–2 µm fraction. These

Physical Fractionation of Organic Matter

317

conditions must be recomputed each time based on Stokes law as a function of the operating conditions (cf. Chap. 2).

–2,000

– – – -

Fig. 9.14. Particle size and centrifuge fractionation after dispersion of the total soil by agitation with glass balls and dispersion of the fraction <50 µm with ultrasounds (Balesdent et al.1991)

The above conditions were calculated by Balesdent et al. (1991) for Stokes diameters of 0.2 or 2 µm, a particle density of 2.5 g cm−3 and the data specific to their equipment. The density used corresponds more to mineral than to organic particles. Thus the fractions indicated do not strictly correspond to the size of organic particles, but it is difficult to separate inorganic and organic matters that are associated in the fractions of clay size. After each decantation of the 0–0.2 or 0.2–2 µm supernatant, suspend the centrifugation pellet in water, agitate for 30 min and centrifuge again. These authors advised four sedimentations at 0–0.2 µm then four at 0.2–2 µm. Centrifuge the 0.2–2 µm suspension for 25 min at 2,900g and recover the centrifugation pellet. Mix the supernatant with the

318

Organic Analysis

previously obtained 0–0.2 µm fractions. Flocculate the suspension by adding a 0.5 g (CaCl2) L−1 solution; store overnight and centrifuge. The centrifugation pellet is the 0–0.2 µm fraction and the supernatant is the final organic water-soluble fraction. Fractionation of the clay size particles can be performed more easily by continuous flow ultra-centrifugation (cf. Chap. 2). The fractions over 50 µm should be dried at 60°C, those below 50 µm should be homogenized, frozen and freeze-dried. They are weighed and then crushed to 50 µm for chemical analyses, especially measurement of their C and N content (cf. Chap. 10).

Fractionation using Resin H + and ultrasounds Decantation method. In this procedure described by Feller et al. (1991), the initial dispersion of the soil is carried out with a cation exchange resin (Amberlite IRN77 in Na+ form) carefully sieved to 500 µm. The sieving operation must be renewed before each fractionation. Split the resin 100-mL portions and place them in polyamide bags (Nytrel TI45) with a mesh size of 45 µm. Place these bags in 60-µm mesh bags (Nytrel TI60). Close the bags with a rubber band. The double bag protects the soil against contamination in the event the bag should break. Dispersion is then carried out by agitation of 20 g air dried soil for 16 h with 300 mL distilled water in a 1 L bottle containing one resin bag. Remove the bag from the suspension, wash abundantly with water and reserve for measurement of the small quantities of 20–50 µm soil fraction that may be trapped in the bag (weigh the fraction remaining on a 20-µm sieve after recovery and wash the resin on a 50-µm sieve). The remaining operations can be performed following the procedure described in “Agitation with Glass Balls and Ultrasonification”. However, the procedure of Feller et al. although very similar to the previous section, includes slight differences in ultrasonic energy, and in the clay fractionation method. Sieve the soil–water suspension to 200 and 50 µm. Wash the material remaining on the mesh of the sieves and subject the 0–50 µm suspension obtained in fractions of 1 L to ultrasounds. The apparatus (250 TH, US Annemasse) uses a frequency of 20 kHz, variable electric output (0 to 300 W), and is equipped with a probe with a flat head 9 mm diameter. This sounding head is located 2.5 cm from the bottom of the suspension, ultrasound is applied continuously for 7 min at 75% of maximum capacity, i.e. 0.23 W mL−1 suspension, approximately 100 J mL− 1 . Sieve the 0–50 µm suspension, wash the material remaining on the mesh, then transfer the 0–20 µm suspension in two sedimentation cylinders and bring to 1 L with distilled water. Shake the cylinders by

Physical Fractionation of Organic Matter

319

turning them upside down and back (30 reversals) and place them on the lab table during sedimentation of the 0–2 µm fraction (cf. Chap. 2) for subsequent pipette sampling of this fraction. Repeat this operation until exhaustion (minimum five times). The sediment remaining at the bottom of the cylinder is the 2–20 µm fraction. Centrifuge the sampled 0–2 µm suspensions for 1 h at 2,500g to separate the centrifugation pellet (0.2– 2 µm) and the 0–0.2 µm supernatant. Repeat this operation twice. Flocculate all the collected supernatants by additions of 2 mL L–1 of saturated SrCl2. Separate the clear supernatant from the centrifugation pellet (0–0.2 µm fraction) by centrifugation. The following fractions are obtained: – by wet sieving, the 200–2000, 50–200 and 20–50 µm fractions – by sedimentation, the 2–20, 0.2–2 and 0–0.2 µm fractions – a water-soluble organic fraction. Depending on the type of soil, or when too energetic dispersion is not desired, the same technique can be used on the 0–50 µm suspension without the ultrasound treatment. Method with sampling of aliquots. This method was described by Gavinelli et al. (1995) and is faster than the preceding one for measurement of the silt and clay fractions. Using a Robinson pipette, remove aliquots from sedimentation cylinders (cf. Chap. 2). Special Procedure for Sandy Soils The procedure of Feller (1979) was developed on sandy soils. It includes low energy mechanical dispersion by agitation of the soil–water suspensions (100 g soil–300 mL water) for 1 h with three glass balls. Sieving with water followed by separation of the fractions as in “Resin H+ and Ultrasounds” enables recovery of the M2000, F2000, M200, F200, M50, F50, OM, W fractions (M: organomineral fraction, F: organic fragments, number: lower limit of particle size of the fraction, OM: organomineral fraction below 50 µm separated by centrifugation, W: water-soluble fraction). The procedure described in “Resin H+ and Ultrasounds” also includes one modification for use with sandy to clayey–sandy soils. The length of agitation of the soil–water–resin suspensions in the bags is reduced to avoid too much deterioration of the plant debris by sands. The procedure is as follows: Place 40 g of air-dried soil in a 1 L bottle with 300 mL distilled water and a 100 mL bag containing “Amberlite IRN77” cation exchange resin in Na+ form (cf. “Resin H+ and Ultrasounds”). Agitate the bottles on a back and forth shaker for 2 h at moderate speed. Separate the fractions

320

Organic Analysis

above 50 µm by sieving. Agitate the 0–50 µm suspension for 14 h with the resin. The remaining operations are identical to the procedure described in “Resin H+ and Ultrasounds” of this chapter. P 9.2.6 . recision of the Fractionation Methods
The precision of the techniques for physical fractionation of organic matters varies considerably with the type of soil and especially with the stage of development of the organic matter. In general, the smaller the quantity of the fraction, the greater the variability. Repeatability increases with the particle size of the fraction. Relative error resulting from fractionation varies in the same way for percentages by weight or the percentage of the carbon of the fraction compared to total carbon. Because of its weak relative weight, the error on the determination of the coarse and light organic fraction is often the most significant. This error also appears to be linked to the method since Monnier et al. (1962) found for four types of soil, variations ranging from +30 to +60% when comparing the funnel method of Jeanson–Luusinang (1960) with the centrifugation method. Oudinot (1985) found for the fraction with a density lower than 1.4, a relative standard deviation of 28% in the case of a calcareous brown soil and 62% in the case of a fersiallitic soil. Feller (1979) also found a coefficient of variation of 63% calculated on 60 replicates of measurements of a coarse organic F2000 fraction separated by sieving at 2 mm and floating in water. Monnier et al. (1962) obtained for two replicates on four types of soil a pooled relative standard deviation of about 2%.
Table 9.1. Error in the precision of the particle size fractionation of a sandy soil from Senegal (Feller 1979; F: organic fragments, M: organomineral fraction, number: lower particle size threshold, OM: organomineral fraction <50 µm, W: water soluble organic fraction), m: carbon percent of the carbon of the sum of the fractions, RSD: relative standard deviation in percent of measurement for ten replicates fraction F2000 F200 M200 F50 M50 OM W m 3.0 21.6 1.2 8.8 5.8 57.6 2.0 RSD(%) 63 11 42 25 28 6 25

Physical Fractionation of Organic Matter

321

Table 9.2. Repeatability of the particle size fractionation of a ferrallitic soil and a vertic soil (Feller et al. 1991). RSD% m and RSD% Ct = relative standard deviation of the mass fraction and the carbon fraction (total carbon of fraction/total carbon) for four replicates
soil type Ferrallitic fraction (µm) 0–2 2–20 20–2000 0–2 2–20 20–2000 RSD% m 10 25 8 4 4 10 RSD% Ct 23 48 48 8 14 30

Vertic

Feller (1979) also measured the error in the percentages of carbon for each fraction (compared with total soil) obtained by wet sieving and decantation (cf. Special procedure for sandy soils). The results (Table 9.1) underline the significance of the error in precision with respect to quantitative studies of soil organic matter and how error varies with the size of the fraction. The study of Feller et al. (1991) also included an evaluation of precision related both to the mass of the fractions and their carbon contents compared to total soil carbon (Table 9.2). The error was shown to depend on the type of soil. The error was lower in the case of a vertic soil than in a ferrallitic soil. 9.3. Conclusion and Outlook Physical fractionation techniques are often used before other studies on soil organic matter. Indeed, they themselves comprise one of the methods of the study of organic matter. No method enables perfect separation of each component of the soil (plant roots, plant fragments, animal fragments, micro-organisms and metabolites of organomineral complexes). The methods described in this chapter seem to be the most suitable for further development. They have been classified under three main functions: extraction of plant roots; extraction of “free organic matter” corresponding to organic fragments that have not completed deteriorated; fractionation of organic matter in particle-size ranges.

322

Organic Analysis

Apparatuses for root-soil separation are based on the principles of elutriation and underwater sieving. Their complexity and the fact that the operations of separation they perform are not exhaustive, led some authors to prefer manual techniques. Density techniques are relatively simple to use. Coupling density with particle size fractionation of the coarse particles and the use of not very aggressive methods of dispersion of the structural aggregates increases reliability. The techniques of particle size fractionation enable more extended classification of organic matter, in particular of the three main organic and organomineral compartments mentioned by Feller (1994). Along with a description of the techniques of physical fractionation, this chapter describes the main types of soil on which the techniques were tested. Adaptations are probably necessary for other soils or to fulfil certain specific research objectives. These adaptations should be also helped by observations of Christensen (2001), Six et al. (2002), Rovira and Vallejo (2003) or Xu et al. (2003). The observations in this chapter should be taken into account, especially precautions related to the use of ultrasounds and dispersing agents; the aim being to obtain better separation of organic fragments and organomineral complexes with less destruction of organic entities.

References
Abo F (1984) Influence du bore et du manganèse sur la nutrition, le développement et la production de blé sur sols de régions tempérée et aride., Thèse d’état, université Paris VII, 390 p Ahmed M and Oades JM (1984) Distribution of organic matter and adenosine triphosphate after fractionation of soils by physical procedures. Soil Biol. Biochem., 16, 465–470 Anderson DW and Paul EA (1984) Organo-mineral complexes and their study by radiocarbon dating. Soil Sci. Soc. Am. J., 48, 298–301 Anderson DW, Saggar S, Bettany JR and Stewart JWB (1981) Particle size fractions and their use in studies of soil organic matter : I. The nature and distribution of forms of carbon, nitrogen and sulfur. Soil Sci. Soc. Am. J., 45, 767–772 Anderson JM and Ingram JSI (1989) Tropical soil biology and fertility (TSBF) : a handbook of methods., C.A.B. International, 171 p Andreux F, Bruckert S, Correa A and Souchier B (1980) Sur une méthode de fractionnement physique et chimique des agrégats des sols : origines possibles de la matière organique des fractions obtenues. C.R. Acad. Sci. Paris, 291, 381–384

Physical Fractionation of Organic Matter

323

Arrouays D (1994) Intérêt du fractionnement densimétrique des matières organiques en vue de la construction d’un modèle bi-compartimental d’évolution des stocks de carbone du sol. Exemple après défrichement et monoculture de maïs grain des sols de “touyas”. C.R. Acad. Sci. Paris, 318, II, 787–793 Balesdent J, Pétraud JP and Feller C (1991) Effets des ultrasons sur la distribution granulométrique des matières organiques des sols. Sci. du Sol., 29, 95–106 Bartoli F, Burtin G and Herbillon AJ (1991) Disaggregation and clay dispersion of oxisols : Na resin, a recommended methodology. Geoderma, 49, 301–317 Beare MH, Hendrix PF and Coleman D.C (1994) Water-stable aggregates and organic matter fractions in conventional-and no-tillage soils. Soil Sci. Soc. Am. J., 58, 777–786 Besnard E, Chenu C, Balesdent J, Puget P and Arrouays D (1996) Fate of particulate organic matter in soil aggregates during cultivation. European Journal of Soil Science, 47, 495–503 Bonzon B and Picard D (1969) Matériel et méthodes pour l’étude de la croissance et du développement en pleine terre des systèmes racinaires. Cah. ORSTOM sér. Biol., 9, 3–9 Bruckert S, Andreux F, Correa A, Ambouta KJM and Souchier B (1978) In Proc. 11e Congrès A.I.S.S., Edmonton, Canada Bruckert S (1994) Analyse des complexes organo-minéraux des sols. In – Pédologie 2. Constituants et propriétés du sol, Bonneau and Souchier ed., 2nd ed., Masson, Paris, 275–295 Cambardella CA, Elliott ET (1993) Carbon and nitrogen distribution in aggregates from cultivated and native grassland soils. Soil Sci. Soc. Am. J., 57, 1071–1076 Catroux G and Schnitzer M (1987) Chemical, Spectroscopic and Biological Characteristics of the organic matter in particle size fractions separated from an aquoll. Soil Sci. Soc. Am. J., 51, 1200–1207 Christensen BT (1992) Physical fractionation of soil and organic matter in primary particle size and density separates. In Advances in soil science, No. 20, Springer Berlin Heidelberg New York Inc, 1–90 Christensen BT (2001) Physical fractionation of soil and structural and functional complexity in organic matter turnover. Eur. J. Soil Sci., 52, 345–353 Dabin B (1976) Méthode d’extraction et de fractionnement des matières humiques du sol. Application à quelques études pédologiques et agronomiques dans les sols tropicaux. Cah. Orstom, Sér. Pédol., XIV, 287–297 Edwards AP and Bremner JM (1967) Dispersion of soil particles by sonic vibration. J. Soil Sci., 18, 47–63 Elliott ET and Cambardella C.A (1991) Physical separation of soil organic matter. Agriculture, ecosystems and environment. Elsevier Science Amsterdam, 34, 407–419

324

Organic Analysis

Elliott ET, Palm CA, Reuss DE and Monz CA (1991) Organic matter contained in soil aggregates from a tropical chronosequence : correction for sand and light fraction. Agriculture, Ecosystems and Environment, 34, 443– 451 Feller C (1979) Une méthode de fractionnement granulométrique de la matière organique du sol. Cah. ORSTOM sér. Pédol., XVII, 339–346 Feller C (1994) La matière organique dans les sols tropicaux à argile 1:1. Recherche de compartiments organiques fonctionnels. Une approche granulométrique., IRD-Orstom, Paris, thèses et documents microfichés Feller C, Burtin G, Gérard B and Balesdent J (1991) Utilisation des résines sodiques et des ultrasons dans le fractionnement granulométrique de la matière organique des sols. Intérêt et limites. Sci. du Sol., 29, 77–93 Ford GW, Greenland DJ and Oades JM (1969) Separation of the light fraction from soils by ultrasonic dispersion in halogenated hydrocarbons containing a surfactant. J. Soil Sci., 20, 291–296 Gavinelli E, Feller C, Larré-Larrouy MC, Bacyé B, Djegui N and Nzila JdD (1995) A routine method to study soil organic matter by particle-size fractionation, examples for tropical soils. Commun. Soil Sci. Plant Anal., 26, 1749–1760 Genrich DA and Bremner JM (1972) A reevaluation of the ultrasonic vibration method of dispersing soils. Soil Sci. Soc. Amer. Proc., 36, 944–947 Golchin A, Oades JM, Skjemstad JO and Clarke P (1994) Study of free and occluded particulate organic matter in soils by solid state 13C CP/MAS NMR spectroscopy and scanning electron microscopy. Austral. J. Soil Res., 32, 285–309 Greenland DJ and Ford GW (1964) Separation of partially humified organic materials from soils by ultrasonic dispersion. Trans. 8th Int. Congr. Soil Sci., 3, 137–148 Gregorich EG, Kachanoski RG and Voroney RP (1988) Ultrasonic dispersion of aggregates : distribution of organic matter in size fractions. Can. J. Soil. Sci., 68, 395–403 Hassink J and Dalenberg JW (1996) Decomposition and transfer of plant residue 14C between size and density fractions in soil. Plant Soil, 179, 159– 169 Hénin S and Turc L (1949) Essai de fractionnement des matières organiques du sol. C.R. Acad. Sci., Paris, 35, 41–43 Jeanson Luusinang C (1960) Fractionnement par densité de la matière organique des sols. Ann. Agron., 11, 481–496 Kilmer VJ and Alexander LT (1949) Methods of making mechanical analyses of soils. Soil Sci., 68, 15–24 Ladd JN, Parsons JW and Amato M (1977) Studies of nitrogen immobilization and mineralization in calcareous soils – I, Distribution of immobilized nitrogen amongst soil fractions of different particle size and density. Soil Biol. Biochem., 9, 309–318 Lein ZJ (1940) Les formes de liaison de l’humus avec la partie minérale des sols. Pochvovedeniye

Physical Fractionation of Organic Matter

325

Mc Gill W.B., Shields J.A. and Paul E.A (1975) Relation between carbon and nitrogen turnover in soil organic fraction of microbial origin. Soil Biol. Biochem., 16, 465–470 Meijboom FW, Hassink J and van Noordwijk M (1995) Density fractionation of soil macroorganic matter using silica suspensions. Soil Biol. Biochem., 27, 1109–1111 Monnier G, Turc L and Jeanson–Luusinang C (1962) Une méthode de fractionnement densimétrique par centrifugation des matières organiques du sol. Ann. Agron, 13, 55–63 Oudinot S (1985) Fractionnement physique de la matière organique. Distribution du carbone natif et marqué entre les fractions granulométriques de deux sols méditerranéens incubés avec du matériel végétal marqué au 14 C., DEA, ENSAIA-INP de Lorraine, France, 30 p Pansu M and Sidi H (1987) Cinétique d’humification et de minéralisation des mélanges sols-résidus végétaux. Sci. du sol, 25, 247–265 Pansu M, Bottner P, Sarmiento L and Metsellaar, K (2004) Comparison of five soil organic matter decomposition models using data from a 14C and 15 N labeling field experiment. Global Biogeochem. Cycles, 18, GB4022, doi: 10.1029/2004GB002230 Pansu M, Gautheyrou J and Loyer JY (2001) Soil Analysis – Sampling, Instrumentation and Quality control., Balkema, Lisse, Abington, Exton, Tokyo, 489 pp Pearson RW and Truog E (1937) Procedure for the mineralogical subdivision of soil separates by means of heavy liquid specific gravity separations. Soil Sci. Soc. Am. Proc., 2, 109–114 Puget P, Besnard E and Chenu C (1996) Une méthode de fractionnement des matières organiques particulaires des sols en fonction de leur localisation dans les agrégats. Comptes Rendus de l’Académie des Sciences, Paris, 322, 965–972 Puget P, Chenu C and Balesdent J (1995) Total and young organic matter distributions in aggregates of silty cultivated soils. Eur. J. Soil Sci., 46, 449–459 Roscoe R, Buurman P and Velthorst EJ (2000) Disruption of soil aggregates by varied amounts of ultrasonic energy in fractionation of organic matter of a clay latosol: carbon, nitrogen and δ13C distribution in particle-size fractions. Eur. J. Soil Sci., 51, 445–454 Rouiller J, Burtin G and Souchier B (1972) La dispersion des sols dans l’analyse granulométrique. Méthode utilisant les résines échangeuses d’ions. Bull. ENSAIA Nancy, France, XIV, 193–205 Rovira P and Vallejo VR (2003) Physical protection and biochemical quality of organic matter in Mediterranean calcareous forest soils: A density fractionation approach. Soil Biol. Biochem., 35, 245–261 Sallih Z and Pansu M (1993) Modelling of soil carbon forms after organic amendment under controlled conditions. Soil Biol. Biochem., 25, 1755– 1762

326

Organic Analysis

Shaymukhametov MS, Titova NA, Travnikova LS and Labenets YM (1984) Use of physisical fractionation methods to characterize soil organic matter. Translated from : Pochvovedeniye, 8, 131–141 Sidi H (1987) Effet de l’apport de matière organique et de gypse sur la stabilité structurale de sols de région méditerranéenne., Thèse Docteur ingénieur, INA Paris Grignon Six J,. Callewaert P, Lenders S, De Gryze S, Morris SJ, Gregorich EG, Paul EA and Paustian K (2002) Measuring and Understanding Carbon Storage in Afforested Soils by Physical Fractionation. Soil Sci. Soc. Am. J., 66, 1981–1987 Six J, Schultz PA, Jastrow JD and Merckx R (1999) Recycling of sodium polytungstate used in soil organic matter studies. Soil Biol. Biochem., 31, 1193–1196 Smucker AJM, McBurney S and Srivastava AK (1982) Separation of roots from compacted soil profiles by the hydropneumatic elutriation system. Agron. J., 74, 500–503 Soil Survey Staff (1960) Soil classification – A comprehensive system, 7th Approximation. USDA, SCS, 265 p Tiessen H and Stewart JWB (1983) Particle-size fractions and their use in studies of soil organic matter: II. Cultivation effects on organic matter composition in soil fractions. Soil Sci. Soc. Am. J., 47, 509–514 Turchenek LW and Oades JM (1979) Fractionnation of organo-mineral complexes by sedimentation and density techniques. Geoderma, 21, 311–343 Watson JR (1971) Ultrasonic vibration as a method of soil dispersion, Soils and Fertilizers, 34, 127–134 Xu YC, Shen QR and Ran W (2003) Content and distribution of organic N in soil and particle size fractions after long-term fertilization. Chemosphere, 50, 739–745

10

Organic and Total C, N (H, O, S) Analysis

10.1 Introduction
10.1.1 Soil Organic Matter Organic matter plays a determining role in pedogenesis and can drastically modify the physical, chemical, and biological properties of soil (structure, plasticity, colour, water retention, CEC, and AEC). The fundamental processes of evolution include phenomena of mineralization and immobilization and, in particular, of carbon and nitrogen. Mineralization allows the transformation of organic residues into inorganic compounds in the soil, the atmosphere, and the hydrosphere, these are then usable by flora and by micro-organisms. Immobilization is the transformation of organic matter into more stable organic and organomineral compounds with high molecular weights that are fixed in the interlayer spaces of clays. These processes are summarized by the following diagram:

Soil organic matter and necromass

Mineralization Immobilisation Gains
Apparent equilibrium

CO2­, CH4­ + NH3­ « NH4 « ­N2O, N2, NOx NO2 «NO3
-

Losses

328

Organic Analysis

This cycle includes phases of mineralization, humification, ammonification, immobilization, nitrification, and volatilization under the action of specific micro-organisms (Pansu et al. 1998) and is influenced by a number of factors, of which the most significant are: – climate (temperature and moisture and their effects on microbial activity, micro-fauna and micro-flora) – topography – types of vegetation and litters – the nature of the parental material (mainly texture, mineralogy and pH) – time (age of the soil and state of equilibrium) and, in addition, for cultivated soils – the effects of farming practices such as ploughing, irrigation, burning, addition of manure, fertilizer, pesticides – types of crops, exports by crops and use of crop residues. Soil C and N contents can vary considerably, e.g. a tropical oxisol can contain less than 2% of total C while in andosols and histisols, total C can exceed 30%. The dynamics of the transformation of C and N are very complex and difficult to model accurately. In addition to the measurement of total C and N, a simple index is needed to clarify the dynamics and allow samples from different climatic zones to be compared, as well as to identify the evolution of the organic matter in a soil profile, or the presence of a buried organic A horizon. The first in situ observations of humified litters (A00) of the organic A0 horizon (which are of varying thickness) and of the eluvial or illuvial horizons provide very important information. The evolution of the organic matter is then analysed by studying its chemical structure and physicochemical properties. The different forms of matter can be separated into characteristic entities of varying degrees of polymerization and quantified by physical separation (cf. Chap. 9) by their resistance to hydrolysis (in acid and basic media) and their solubility in specific solvents and reagents (cf Chaps. 11 and 12). Laboratory techniques such as gel permeation chromatography allow determination of the molecular weights of the humified substances after purification. UV, visible, or IR absorption and other spectrographic methods allow identification of the molecular structures and of rates of polymerization. Both active functional groups and the formation of the clay–humus complexes can be studied (cf. Chap. 12). The elementary analysis described in this chapter enables the total chemical composition of the organic matters to be established. The different factors that control humification, especially climate, parental material, and biomass, result in physicochemical constraints that,

C, N (H, O, S)

329

in turn, result in a given type of humus e.g. Mor, Moder, forest and calcic Mulls, Anmoor or peats. Microscopic observation (cf. Chap. 8) of the morphology of the systems at different scales allows characterization of the interfaces of organic and inorganic matter and of the mechanisms of humification. Kinetic methods can also be used to analyse the biogeochemical dynamics of the organic residues: – By measuring, for example, the CO2 released per unit of time by means of portable chromatographs or IR captors installed in situ (respirometry, biological evolution). – By using 14C, 13C, and 15N isotopic tracers to monitor the transformation of organic substances added to the soil (turnover rate of soil organic matter). – By studying the differentiation of stable isotopes such as 13C, 14C, or 15N and and isotopic ratios for studies of paleoclimatology and geochronology.
Table 10.1. Typical C:N ratios of a few main types of humus type calcic eutrophic Mull forest Mull moder mor
hydromorphy calcic eutrophic peats acid oligotrophic peats anmoor

C:N ≅ 10 12–15 15–25 > 20 < 30 ≅ 40 variable

pH range 7.0–8.4 5.5–6.5 4.0–5.0 < 4.0 – – –

The majority of these methods require a high degree of specificity and highly sensitive sensors because of the scales of measurement needed to measure extremely weak variations. They are time consuming and expensive and are not sufficiently universal for use in serial analysis. On the other hand, total-C and -N can be measured using simple methods that are accessible to all laboratories. Improvement in equipment for dry analysis (e.g. CHN(OS)) now makes it possible to standardize analyses and combine precision, speed, and automation. Some of this equipment can handle representative samples of more than 100 mg. The C:N ratio of the soil in the surface horizons can be determined only using information on total C and N (Table 10.1). This information can then be used as an index that provides relatively reliable information on the biological activity and equilibrium of the two elements that have been subjected to the antagonistic processes of mineralization and immobilization. In regions with a temperate climate, the C:N ratio is

330

Organic Analysis

about 10–12 for uncultivated soils and generally decreases with an increase in soil depth. In certain soils N can be significantly occluded in clays, especially in deep horizons. In forest soils, peat horizons, or podzols, C:N ratios can reach 20–30 or even higher because of the formation of only slightly biodegradable complexes which are low in nitrogen (e.g. Spodic horizons). At C:N ratios below a threshold near 20, positive N net mineralization is generally observed. In cultivated soils, farming residues recycled in the field have C:N ratios ranging between 15 and 60 due to the presence of lignin-cellulose compounds with a slow rate of degradation. Under forest with acidifying litter, the C:N ratios can reach 150 or even higher. 10.1.2 Sampling, Preparation of the Samples, Analytical Significance Equipment

– strainer with 2 mm round holes, AFNOR NF34. – cutting grinder equipped with a 125 µm mesh sieve (AFNOR NF22) and a watertight collector (for litter). – grinder with retractable hammer equipped with an AFNOR NF22 sieve and a watertight collector (mineral or organo-mineral horizons). – agate mortar and pestle. – analytical balances (±0.1 mg or ±0.01 mg depending on test specimens). – drying oven regulated at 105°C.
Procedures and Precautions As the heterogeneity of the soil surface horizons near the litters is very high, sampling is difficult and must be carried out with great care. The way samples are collected will subsequently affect the validity of the results. Drying should be carried out in contact with air in a wellventilated room. At the laboratory, samples should be crushed to 2 mm to separate nondecayed or only slightly decayed plant debris. Care should be taken to avoid breaking up the organic fragments that have retained their original texture as this could overload the sample significantly. This stage affects the significance of the analytical result as does the weight of the test specimens during analysis (Pansu et al. 2001).

C, N (H, O, S)

331

Drying increases the fixing of ammonium particularly if the parental matrix contains 2:1 clays such as montmorillonite or vermiculites. The action of the micro-organisms may not have stopped during storage depending on soil respiration, lignin content, or on residual moisture in the air-dried sample (in andosols and histisols, the moisture rate of air dried soil can still be 60% after 6 months). Grinding a whole 2-mm particle-size sub-sample to 125 µm particle size (AFNOR NF22 sieve) can modify the moisture and equilibrium of the reactive surfaces. Moisture content is used to correct the analytical results and must be measured at the same time as analytical sampling is performed. If drying at 105°C does not exceed 3 or 4 h, there is generally no significant loss of C and N in gaseous form; but drying can slightly increase fixing of N in the clay lattice. In the case of instrumental analysis (CHN(OS)), drying the samples at 105°C can avoid the need to correct the results and can limit clogging of the water traps. Careful grinding into fine particles that are homogenous in size is necessary to improve the reproducibility of the results, soil powder sampling can vary between 5 and 500 mg depending on the instrument used for dry analyses (CHN(OS)). In the case of wet analyses, grinding ensures a more regular and complete attack by considerably increasing the solid–liquid interfaces. Expression of the Results If the analytical results are to be used for agronomic purposes, it is advisable to take the soil density into account, particularly in the case of peats, andosols, and histisols where apparent density can approach 0.30. In this case, the concentration expressed per mass unit may be far from the inorganic contents actually available for the plants per unit volume and it is thus essential to correct for density. In the case of soils with high contents of gravels, stones, rocks, or nondecayed plant debris that display a significant rejection rate during preparation of 2 mm soil samples, it is also better to correct the rough results of analysis to obtain a value approaching the quality of the soil per unit of volume. Preliminary Tests A good knowledge of the formation processes and agricultural use of the soils concerned makes it possible to limit the number of tests required.

332

Organic Analysis

The number of tests also depends on the analytical method to be used (Table 10.2). It should be noted that by convention, the “total organic matter of the soil” corresponds to the transformed organic forms and excludes intact plant and animal residues. In practice, since the separation of light or non-decayed fragments of organic matter (cf. Chap. 9) is difficult in repetitive analysis, particles of a size lower than 2 mm are considered to be an essential part of the sample. Living micro-organisms are integrated. A range of different types of tests can be carried out to obtain more detailed knowledge on the analytical substrate to be in a position to choose the appropriate analytical procedures. Examination with a magnifying glass enables confirmation of presence of seashells, limestone amendments, coals, etc. (soils under crops, coastal soils, calcareous soils, etc.). HCl Tests: below pH 7.4–7.0, indicates the presence of carbonates and bicarbonates only in the form of isolated particles.
Table 10.2. Analytical methods method type
combustion < 360°C dry methods combustion > 360°C

classical N Kjeldahl method wet methods

cold C oxidation
hot C oxidation

source of interference gypsum, losses of water above 150°C various forms of water (hydration) carbonate decomposition CO2 Na2CO3 melting (clogging) various forms of water (bound water) NO3–, NO2– random recovery (not quantified if not previously reduced) NH4+ fixed in the crystal lattice of 2:1 clays (random recovery) results are too low (multiplicative factor) presence of chloride, oxidative and reducing agents presence of chloride, oxidative and reducing agents

NO3–, NO2– test: a rapid test using a soil analytical kit is useful in intensively cultivated and waterlogged soils. Fixed NH4 test: only used in soils containing 2:1 clays to check ammonium concentration with two different methods and possibly by destruction of the crystal lattice (cf. Sect. 28.3.5 of Chap. 28).

C, N (H, O, S)

333

Preliminary Destruction of Carbonates and Bicarbonates (Dry Combustion Methods) The samples should be treated at room temperature with a 0.1 mol (HCl) L–1 solution until the end of the reaction. If the presence of dolomite, siderite, or biogenic calcite is suspected, contact time should be increased to 2 or 3 h. CaCO3 + 2H+ → Ca2+ + CO2↑ + H2O Samples containing siderite can be treated using the hot H3PO4 or CH3COOH 0.3 mol (H+) L–1 for 5 h. Destruction is difficult and often incomplete. Dry carefully. Total C – inorganic C = organic C

(10.1)

Care should be taken to not lose soluble organic C in the acids (e.g. from amino-acids or phospholipids). Formation of hygroscopic salt can disturb weighing. The presence of manganese dioxide in the soil can cause the release of Cl2 starting from HCl. In arid or semi-arid regions, soluble salts that may be present (e.g. carbonates, chlorides, or sulphur compounds) can slow down the organic oxidation of C because of their low melting points (e.g. sodium carbonate) and consequently disturb measurements by dry combustion.

10.2 Wet Methods
10.2.1 Total Carbon: General Information Strictly speaking the “total carbon” of the soil comes from two principal sources (10.1): – Organic carbon (only slightly processed organic residues of plant and animal origin, humus, charcoal, fossil organic matter, microorganisms). – Inorganic carbon possibly present in the form of carbonates and bicarbonates. In the majority of methods, the gas phases present in the atmosphere of the soil (CO2 linked with biological activity, CH4) are not taken into account. Some ambiguity persists in the terminology and methods used. Measurements carried out on non-processed soil samples (without preliminary elimination of carbonates) using dry combustion methods in

334

Organic Analysis

CHN(OS) apparatuses give organic and inorganic forms of “total carbon”. Measurements by oxydo-reduction using wet oxidation give only “organic carbon” corresponding to humified forms and organic matter of debris that are still rich in non-transformed cellulose, but does not include charcoal or fossil organic matter. Although wet methods at room temperature do not allow complete attack of the humus (a correction index will be needed), they are nevertheless used for the determination of “total organic carbon”. “Total inorganic carbon” can be measured using the methods described in Chap. 17 but with varying precision due to the slow chemical decomposition of magnesium carbonate (MgCO3) and especially of siderite (FeCO3). The term “total organic matter” is often used. Empirically it expresses “total organic carbon” determined by oxidation–reduction but corrected by a coefficient based on the assumption that organic matter contains mainly humic acids at approximately 58% of carbon (100/58 = 1.724 van Bemmelen factor). In fact this rate is far from constant even for the horizons near the soil surface. The coefficient is thus not very realistic, particularly for soils containing not very humified Mor or Moder, forest soils, and peats. In these cases a coefficient of up to 2 or even 2.5 will be required. The term “total organic matter” is thus an estimation and cannot be used as an index. In practice, the term “total carbon” is incorrectly used by some authors to indicate “total organic carbon” as well as to establish balances in total analysis, C:N ratios, or to compare the total C contents of the different horizons of a soil profile and the organic distribution of C as a function of depth. The term total organic carbon should cover all the organic substances resulting from the humification of C in the soil (microbial residues, humic substances) under the influence of biochemical and chemical reactions. Additionally it represents light organic matter still not completely decomposed that could not be separated on a sieve during the preparation of the samples (litters, coarse organic plant or animal fragments under 2 mm in size). Fossil organic matter (coal, naphthas, resins, etc.) and charcoals in regions exposed to forest fires or in regions with slash and burn agriculture are not subject to this type of dynamics, which means they do not have to be taken into account when wet redox methods are used. However they are always included when dry combustion methods are used and this can lead to difficulties when the results obtained by the two methods have to be compared. Thus the study of the total organic carbon stock may need to be refined:

C, N (H, O, S)

335

– By micro-morphologic observations at different scales to determine the relative proportions of the contents of unprocessed and humified organic matter together with selective extractions in different mediums. – By the determination of the origin, the nature and the rates of mineralization of the different forms as a function of the pH, the nature of the clays and clay–humus complex, and of soil management practices. Wet methods require only relatively inexpensive equipment, which means they can be used in all laboratories. These methods make it possible to work with big samples which are more representative of the natural environment. On the other hand, they are time consuming, require the handling of very corrosive products and the elimination of polluting products (Cr3+, H2SO4), which can pose problems for the environment.

10.2.2 Organic Carbon by Wet Oxidation at the Temperature of Reaction
Introduction The determination of total organic carbon by oxidation with potassium dichromate in a strong acid open medium, was proposed first by Schollenberger (1927) then by Walkley and Black (1934) from which it takes its name. After a stage of oxidation/mineralization at the temperature of reaction for a given length of time, the non-reduced dichromate in excess is back titrated by ferrous iron. Many authors have studied the factors that affect C mineralization: acid concentration (H2SO4, H3PO4), potassium dichromate concentration, oxidation temperature (from the temperature of reaction to +210°C), the time of contact, the need to condense the vapour to avoid too high concentration of the medium, and to limit destruction of the oxidant by avoiding overheating of the walls. The choice of the temperature resulted in two different types of methods: – At the temperature of reaction ≅ 120°C (Walkley and Black 1934). – At standardized boiling ≅ 150°C (Anne 1945; Mebius 1960). Different procedures were proposed for back titration of dichromate excess such as soil/solution separation by centrifugation and filtration, but direct volumetric titration in the soil suspension was the most widely adopted. Redox indicators (diphenylamine, barium diphenylamine sulphonate, N-phenyl anthranilic acid, ferrous O-phenanthroline) can be absorbed on clays. Additives (e.g. NaF, H3PO4) allow better reading by sequestering the coloured products that are formed or dissolved (e.g. Fe3+) and which can mask the reaction.

336

Organic Analysis

Principle Mineralization Organic forms of C are oxidized in the presence of excess dichromate. The reaction in a concentrated acid medium is exothermic (≅ 120°C). It develops at a fast kinetics under the following conditions: 120°C ⎯ 3C + 2Cr2O72– + 16H+ ⎯⎯⎯ → 4Cr3+ + 8H2O + 3CO2 The amount of reduced dichromate is considered to be quantitatively linked to the organic C content of the sample. The likelihood of reduction is assumed to be identical for different forms of organic C and the reducing power is assumed to be constant during mineralization. In practice, at the temperature of reaction without heating, a factor of correction will be required because only the most active forms are oxidized, i.e. 60%–80% of organic matter. This factor was fixed at 1.30 (100/76) to take the variable reactivity of the organic forms into account, but can vary between 1.10 and 1.45 depending on the soils and on the types of vegetation. The inorganic forms (carbonates, bicarbonates) are destroyed and do not play any role except in the consumption of acid and the production of foam. The precipitation of calcium sulphate can be problematic during final titration if a spectro-colorimetric method is used. CaCO3 + H2SO4 → CaSO4↓ + H2O + CO2↑ Titration Volumetric back titration of the Cr+VI dichromate not consumed by organic C is carried out by reduction (ferrous sulphate or Mohr’s salt) in the presence of an indicator. Cr2O72– + 6Fe2+ + 14H+ → 2Cr3+ + 6Fe3+ + 7H2O Sodium fluoride or phosphoric acid (H3PO4) can be added to fix the ferric iron that is formed or dissolved, and to improve the detection of the equilibrium point of titration: Fe2O3 + 3H2SO4 → Fe2(SO4)3 + 3H2O Fe3+ + 6F– → FeF63– (uncoloured) Nevertheless, the addition of fluoride can result in the formation of hydrofluoric acid which attacks glass and silicates: 2NaF + H2SO4 → Na2SO4 + 2HF SiO2 + 6HF → H2SiF6 + 2H2O

C, N (H, O, S)

337

It is thus necessary to clean the lab glassware immediately after titration and to reserve this glassware for the determination of C using this redox method. Interferences In saline soils, chlorides cause a positive error by forming chromyl chloride: K2Cr2O7 + 6H2SO4 + 4KCl → 2CrO2Cl2 + 6KHSO4 + 3H2O Ferrous iron that may be present is oxidized by dichromate and thus modifies the quantity of ferrous sulphate necessary for back titration. Corrections will be necessary in waterlogged soils in which Fe2+ is sometimes abundant. The method cannot be used in acid sulphated soils that are rich in pyrite (FeS2): 5Cr+VI + FeS2 → Fe3+ + 2S+VI + 5Cr3+ A high level of Mn2+ can also interfere, as can iron metal that can results from wear of the grinding equipment. Equipment – Analytical balance (±1/10 mg) and a top-loading balance with a capacity of 500 g (±1 mg). – 500 mL wide-neck Pyrex Erlenmeyer flasks. – Insulating plates. – Teflon flask dispenser. – Burette for titration. – Magnetic stirrer with Teflon bars. Reagents All the reagents should be of analytical reference grade: – Distilled or bi-distilled water (avoid water that has been deionised by ion exchange as it can contain fine particles of ion-exchange resins). – 1 mol (e–) L–1 potassium dichromate (standard): in a 1 L volumetric flask, dissolve 49.040 g of K2Cr2O7 (dried under vacuum or on P2O5 in a desiccator) in 800 mL of bi-distilled water, then adjust to 1,000 mL. – Concentrated sulphuric acid, H2SO4 d = 1.84. – 0.5 mol (H+) L–1 sulphuric acid solution: in a 1,000 mL Pyrex volumetric cylinder, add 800 mL distilled water, then slowly add 13.9 mL of concentrated sulphuric acid, homogenize. Allow to cool and bring to 1 L with distilled water.

338

Organic Analysis

– 0.5 mol (e–) L–1 iron and ammonium sulphate (Mohr’s salt): in a 1,000 mL volumetric flask, dissolve 196.05 g of Fe(NH4)2SO4,6H2O (dried on P2O5 in a desiccator) in approximately 800 mL of 0.5 mol (H+) L–1 H2SO4 solution. Adjust to 1,000 mL with the 0.5 mol (H+) L–1 H2SO4 solution. The liquid should be clear and pale green in colour. – Concentrated phosphoric acid H3PO4 d = 1.71 (85%). – Sodium fluoride NaF in powder form. – Diphenylamine solution: dissolve 0.5 g diphenylamine in 100 mL of concentrated H2SO4. Pour into 20 mL water and store in a brown glass bottle with a ground stopper with dropping pipette. Other indicators can be used:
N N

H N

which forms a – o-Phenanthroline (1–10 phenanthroline) ferroine complex with Fe2+: dissolve 14.85 g o-phenanthroline monohydrate and 6.95 g ferrous sulphate (FeSO4, 7H2O) in 800 mL distilled water. Bring to 1,000 mL in a volumetric flask and store in a brown glass bottle. – Barium diphenylamine sulphonate Ba(C6H5–NH–C6H4–SO3)2. Dissolve in distilled water.
COOH NH

– N phenyl anthranilic acid Procedure

.

If total N was analysed beforehand, determine the approximate weight of soil required to obtain a sample specimen containing between 10 and 25 mg C (Table 10.3). Weigh this sample specimen (± 0.1 mg), transfer it in a wide-neck 500 mL Erlenmeyer flask and add exactly 10 mL of the potassium dichromate 1 mol (e–) L–1 solution. Homogenize carefully to avoid making the suspension go up the walls of the flask. Quickly add 20 mL of concentrated sulphuric acid with a Teflon dispenser. Agitate by rotation for one minute to homogenize (the temperature of reaction is approximately 120°C). Place on an insulating plate and let oxidation take continue for 30 min. Add 200 mL distilled water, then 10 mL of phosphoric acid (or approximately 5 g of sodium fluoride with a suitable spatula). Homogenize. Add three drops of diphenylamine.

C, N (H, O, S)

339

Titrate the excess dichromate with the 0.5 mol (e–) L–1 ferrous iron solution (this reagent should be freshly titrated each day). The end of titration is indicated by the change in colour from purplish blue to a rather luminous greenish blue. Determination of the end point is facilitated by adding 1–2 drops of indicator as soon as the colour begins to change.

Note: the solution should still be orange after the attack of the organic matter indicating an excess of dichromate. If the solution is green, start again using a smaller sample of soil.
Table 10.3. Recommended size of test specimen (P g of soil sample) as a function of nitrogen content (on the basis of a C:N ratio of 10). N g kg
–1

sample Pg
10 5 2.5

mg C sample (theoretical)
10–20 15–20 12.5–22.5

N g kg–1
1–2 3–4 5–10

sample Pg
1 0.5 0.2

0.1–0.2 0.3–0.4 0.5–0.9

mg C sample (theoretical) 10–20 15–20 10–20

Expression of the Result It is an accepted fact that the oxygen consumed is proportional to the carbon titrated on the theoretical basis of 1 mL of 1 mol (e–) L–1 dichromate solution oxidizing 3 mg C, i.e. corrected by the attack coefficient (1.3 = 100/76) = 3.9 mg C. This attack coefficient can be modulated as a function of the form of C and by comparison with measurements made by dry combustion: – Total organic C g kg–1 of 105°C dried soil = 3.9(10 – 0.5 V) P where P is the sample mass in g, V is the volume (mL) of Fe2+ solution at a concentration of 0.5 mol L–1 (replace 0.5 by the exact concentration if it is not exactly 0.5) and the quantity of dichromate solution added is 10 mL. – Total organic matter g kg–1 = total organic C × 1.724

/

Control Each day, make two measurements with 10 mL 1 mol (e–) L–1 dichromate solution to check the exact concentration of the Fe2+ solution.

340

Organic Analysis

In each series, carry out two blank titrations with quartz prefired at 1,100°C, and two titration controls on samples of a reference soil of the same type as the soils being analyzed.

10.2.3 Organic Carbon by Wet Oxidation at Controlled Temperature
Introduction and Principle When measurements are made using wet oxidation at the temperature of reaction, the mineralization/oxidation of active organic carbon is always incomplete. The use of a “standardized” corrective factor of 1.3 introduces a variable that is not easily controllable because in practice, this coefficient ranges from 1.10 to 1.40. To mitigate this problem, Anne (1945) then Mebius (1960) proposed carrying out total mineralization by maintaining the sample at a constant temperature throughout the process of mineralization without causing thermal decomposition of the dichromate. Effectiveness and reproducibility were tested between 130 and 210°C with different times of oxidation and different acid/dichromate ratios. Above 150°C the dichromate tends to decompose more and more quickly, thus necessitating relatively short attack times. Strict respect of the procedure and precise control of the attack times and the temperature enable an acceptable level of accuracy. In principles this type of measurement and possible interferences are the same as for the method described in Sect. 10.2.2 earlier. Equipment – A mineralization block (with from 20 to 40 places regulated thermostatically at 150°C (Tecator, Skalar, Technicon, etc.); mechanization is possible in laboratories that carry out many repetitive analyses; in the Skalar system, for example, a sample holder is capped by a device with 20 reflux condensers. After introducing the samples and reagents, place the unit in the programmed heating block; after the period of mineralization, remove the unit and cool, separate the rack condensers; the sample holder advances on rails towards the dilution stage and possibly towards the titration system. – Analytical balance (±1/10 mg). – Digestion tubes with ground joint for the condenser. – Titration burette.

C, N (H, O, S)

341

– Precision volume dispenser (±1/10 mL) with Pyrex glass syringe and Teflon piston. – Magnetic stirrer with 15 mm Teflon bars. Reagents – cf. “Reagents” under Sect. 10.2.2. – 0.2 mol (Fe2+) L–1 solution: weigh 78.5 g of Mohr salt (dried in P2O5 desiccator), dissolve in 800 mL of 0.5 mol (H+) L–1 sulphuric acid solution. Bring to 1,000 mL with the sulphuric acid solution. Procedure Weigh (±0.1 mg) between 200 mg and 2 g of soil (ground to 125 µm particle size and dried on P2O5) to have a sample containing approximately 15 mg C (Table 10.3). In a Pyrex tube (with ground joint for the condenser), add 10 mL of 1 mol (e–) L–1 potassium dichromate solution and 15 mL of concentrated sulphuric acid. Homogenize. Place the tube in a heating block regulated at 150°C and adjust the condenser. The attack should be maintained for 30 min at the same temperature. Leave to cool in the air then transfer in a 250 mL wide-necked Erlenmeyer flask and bring to approximately 100 mL with washing water. Titrate with the 0.2 mol (Fe2+) L–1 solution (or 0.5 mol L–1) in the presence of the indicator and a sequestering agent (cf. “Procedure” under Sect. 10.2.2). The dark purple colour will change to luminous green at the titration point. Controls and Calculation – Titration of the Fe2+ solution (two replicates). – Blank titration of prefired quartz by heating under the same conditions to correct the thermal destruction of the reagent and to establish the effective dichromate concentration (two replicates). – Analyse a standard carbohydrate under the same conditions to check that oxidation is complete and that a correction coefficient will not be needed. Without a correction coefficient, total organic C g kg–1 of 105°C dried 0 . soil = 3(10 –5 V ) P See “Expression of the Result” under Sect. 10.2.2 for explanation of symbols and numbers.

/

342

Organic Analysis

10.2.4 Organic Carbon by Wet Oxidation and Spectrocolorimetry
The French standard NF X31-109 (1993) was published for the determination of organic carbon by sulfochromic oxidation allowing the calculation of the organic matter by means of a multiplying coefficient for use in agronomic studies. Oxidation is conducted at 135°C in a thermostated heating block. The oxidation of a carbon atom requires the transfer of four electrons. Glucose is used as the standard substance and the final determination is by absorption spectrometry at 585 nm on aliquots centrifuged for 10 min at 2,000 g and filtered to eliminate the suspended particles. The method cannot be used in the presence of mineral reducing materials (e.g. Cl–, Fe2+) or of pollution by organic compounds. This standard NF X31-109 (1993) is referred to in the detailed procedure. Notes: – After centrifugation, titration of organic C in the medium can be carried out at 590 or 625 nm (depending on the author with a standard sucrose). – To avoid transfers, a probe spectrometer with optical fibre can be used directly in the clarified medium (Baker 1976).

10.2.5 Total Nitrogen by Wet Method: Introduction
After carbon, hydrogen, and oxygen, nitrogen is the most abundant element in living tissue. It plays a major role in agriculture, nitrogen being an essential element for plant growth. In the soil, the organic forms can reach approximately 90% of total nitrogen. Quantitatively speaking, the total nitrogen value expresses not only the N compounds of the organic matter of the soil and biomass (cf. Chap. 14), but also inorganic nitrogen compounds (cf. Chap. 28). All these compounds represent both short- and long-term reserves, i.e. nitrogen that is directly or potentially available for plants enabling an improvement in yield. On the other hand, total nitrogen cannot quantify the values of transfer of the different forms of nitrogen between living organisms and inorganic materials. Thus total N cannot qualitatively express the diversity of the forms of nitrogen that vary considerably in soils subjected to specific climatic constraints, environmental conditions such as types of vegetation, or different farming systems, nor can it express the complex interactions between micro organizations which control the nitrogen cycle. The components of proteins, carbohydrates, hemicelluloses, cellulose,

C, N (H, O, S)

343

and lignin from living organisms are degraded in the soil, lignin being the most stable fraction. The phases of N immobilization occur in the form of organic and fixed N, and occluded or exchangeable forms of N–NH4+. The phases of nitrification/denitrification produce NO2_, NO3_, N2, nitrogen oxides (with uptake by plant roots and losses through drainage). N processes in the soil can be modified by symbiotic associations between plants (e.g. leguminous plants) and bacteria (rhizobia, actinomycetes) which involve the formation of nodules that enable nitrogen to be fixed from the atmosphere. The analysis of “Total N” using a wet method derives from that proposed by Kjeldahl (1883). Without time-consuming complementary treatments that method does not allow all the forms of N pools to be recovered entirely. The organic and inorganic nitrogen compounds include (in varying proportions): – Inorganic forms, (1) NH4+–N, which is exchangeable or fixed in the mineral or organomineral lattices, (2) NO3–N, which is abundant under intensive cultivation on heavily fertilized soils, (3) NO2–N, which is generally negligible except in waterlogged soils or when it results from polluting wastes; NO3––N and NO2––N are very soluble and are consequently easily leached by water infiltration and run-off. – Entities whose chemical, physical, and biochemical behaviours are well enough defined to enable them to be grouped in selective pools: plant fractions in different stages of decay, active microbial biomass, biological forms (amino acids, amino sugars, proteins of bacterial cells), N subjected to hydrolysis in an acid medium, N of doubtful composition not subjected to hydrolysis. Part of N is also included in complex humified compounds of varying degrees of stability: relatively instable fulvic acids, humic acids with varying degrees of polycondensation, humins (bound on clays and cementing Fe–Al agents) especially protein forms that are weakly attacked by proteases. The evolution of soil organic nitrogen is linked to the molecular forms of humic compounds. The processes of condensation of humic molecules and of formation of organo-mineral compounds modify the stability of the different pools. These pools can be ranked on the basis of increasing stability as follows: fulvic acids < organo-aluminous compounds < organoferric compounds < humic acids < various organo-mineral compounds < humins fixed on clays. The production of ammonium is an indication of instability. At the physical level, the organic layers adsorbed superficially on the mineral or organomineral matrices react more easily than those fixed in the lattices. At the chemical level, the short nitrogen chains are hydrolysed more rapidly than the long chains or the N compounds fixed in clays.

344

Organic Analysis

The addition of water before analysis releases a varying proportion of fixed N by causing swelling of the lattices of certain 2:1 clays. This fixed or occluded N can play a role in plant nutrition (Mengel and Scherer 1981; Keerrthisinghe et al. 1984). Classical Kjeldahl analysis makes it possible to quantify only one part of it; so to control this variable, the mineral matrix must first be destroyed by a mixture of HF–HCl (cf. Sect. 10.3.5 in Chap.10). The question of whether it is possible to fully describe the relative availability of N compounds in the soil by means of models of the chemical and physical compartment that distinguish all the active and passive forms has not yet been answered. But whether the answer is yes or no, the analysis of total organic and inorganic N is an essential component of mathematical models based on a dynamic simulation of the forms of N in the soil – plant – climate systems.

10.2.6 Total Nitrogen by Kjeldahl Method and Titrimetry
Principle Mineralization The aim is to transform organic N forms into ammonium-N form using a wet method in a concentrated sulphuric acid medium in the presence of catalysts.
catalyst → (R3)N + H2SO4 ⎯⎯⎯ (NH4)2SO4 + H2O +CO2 + 2NH4 + H2SO4 → (NH4)2SO4 + 2H+

All nitrogen in amide, imide, nitro N–N, nitroso N–O, or other forms is transformed into the ammonium salt form. The thermal stability and the rise in temperature of the reaction medium are ensured by the addition of K2SO4. Nitrates and nitrites are probably not accounted for, even with a mercury catalyst: 2HNO3 + 6Hg + 3H2SO4 → 3Hg2SO4 + 4H2O + 2NO↑

C, N (H, O, S)
COOH OH

345

Nitrates and nitrites can be reduced by salicylic acid sodium hyposulfite Na2S2O3. NH4+ titration

and

The NH4+–N produced is transferred to a basic medium by steam distillation, and then titrated volumetrically in the presence of an indicator. 2NH4OH + H2SO4 → (NH4)2SO4 + 2H2O (NH4)2SO4 + 2NaOH → Na2SO4 + 2NH3↑ + 2H2O After transfer the NH3 is collected in boric acid (Winkler 1913) and titrated by acidimetry: 4H3BO3 + 2NH4OH → (NH4)2B4O7 + 7H2O (NH4)2B4O7 + 3H2O ⇔ 2H3BO3 + 2NH4BO2 2NH4BO2 + 4H2O ⇔ 2H3BO3 + 2NH4OH It is also possible to delay titration, and first to complete distillation, and then to perform titration. Equipment – Analytical balance (±0.1 mg). – Rack for attacks with gas heating or thermostatic heating blocks regulated at 360°C. – 350 mL Pyrex Kjeldahl flasks, or Pyrex cylinders for heating blocks. – Dosing spatula for catalysts (≅ 500 mg, 1 g, 2 g). – Glass balls of 6 mm dia. – Teflon flask dispenser. – Pyrex ball jacks (Fig 10.1a). – Fume hood with outlets for heavy vapours. – Distillation or steam distillation apparatus. – 250 mL Pyrex Erlenmeyer flasks. – 50 mL burette (±1/10 mL). – Magnetic stirrer with Teflon magnetic stirring bars. – 5,000 g centrifuge and 100 mL Pyrex centrifugation tubes. – Spectrocolorimeter.

346

Organic Analysis

Reagents – Sulphuric acid, H2SO4 d = 1.83. – Sulphuric acid containing 50 g L–1 salicylic acid. – Catalyst: grind and sieve (Afnor NF22) 100 g of potassium sulphate (K2SO4), 20 g of copper sulphate (Cu(SO4),5H2O), 2 g of grey selenium powder (Se); homogenize and store in a wide-necked bottle. – 2% boric acid (H3BO3) solution in distilled water. – Taschiro indicator (Ma Zuazaga 1942): mix one part of 0.1% methyl red ethanol solution with three parts of 0.1% bromocresol green ethanol solution; store in a brown bottle with dropper. – Sodium hydroxide (NaOH) solution ≅10 mol L–1: weigh 2.5 kg of NaOH pellets and carefully dissolve in 6 L of distilled water; let cool in a closed Pyrex bottle (CO2 is eliminated by precipitation of Na2CO3); decant Pyrex bottle then discard the bottom part; store in an airtight bottle. – 1% phenolphtaleine solution in 30% ethanol solution. – Standard solutions of 0.1 and 0.05 mol (H+) L–1 sulphuric acid. – Standard ammonium sulphate solution: weigh 4.714 g of (NH4)2SO4 dried on P2O5; dissolve in deionised water and complete to 1 L; 1mL solution = 1 mg N. Procedure (Macro-Method) Mineralization Weigh 2 g (±0.1 mg) of soil (ground to 125 µm particle size) on nongummed cigarette paper (without nitrogen). Close carefully by twisting the paper. Transfer the sample to a 350 mL glass Pyrex flask. Wet the soil with a jet of distilled water from a wash bottle. Agitate gently until complete homogenisation of the soils. Leave in contact overnight. Carefully (especially with calcareous soils) add 20 mL of concentrated sulphuric acid. With a dosing spatula, add 2 g of catalyst. Add 3 glass balls (diameter 6 mm), place a ball jack in the neck of the flask and place the flask on the rack (Fig. 10.1a). Adjust the gas flame to low and check for the formation of foam. The flask should be tilted at an angle of 45– 60°C to limit the risk of projections and to allow better recovery of condensation on the lower walls (Fig. 10.1a). When the soil organic matter is broken up, the colour of the sample will have faded; raise the heat and boil for 2–3 h without going to dry.

C, N (H, O, S)

347

Distillation and titration by acidimetry Leave to cool. Rinse the jack and the flask walls with about 10 mL distilled water. Fit the flask in the distillation apparatus (Fig. 10.1b). Add 100 mL of 10 mol L–1 sodium hydroxide solution. Turn off the tap to begin steam distillation. Collect the distillate in the Pyrex Erlenmeyer flask containing 20 mL of 2% boric acid solution and 3 drops of Taschiro indicator; take care that the end of the exit tube is below the surface of the liquid in the Erlenmeyer flask. Approximately 80 mL of the distillate are needed for quantitative recovery. Titrate by volumetry with the 0.1 or 0.05 mol (H+) L–1 sulphuric acid solution depending on the estimated quantity of the contents. The end point is indicated by a change in colour from green to greyish purple.

Fig. 10.1. Minimal equipment needed for Kjeldahl titration of total- N, ( a) acid mineralization with device to limit acid loss, (b) distillation of resulting ammonia by addition of soda; for a more powerful apparatus for steam distillation, see Fig. 10.3 in Chap. 14

Expression of the Results The results TN are expressed in mg of nitrogen N per kg of soil dried at 105°C: 1 mL H2SO4 0.1 mol(H+)L–1 ⇔ 0.1 mmol NH3 is 1.4 mg N.

348

Organic Analysis

If VA is the volume (mL) of the 0.1 mol (H+) L–1 H2SO4 solution and P the mass (g) of the test specimen of soil dried at 105°C, the N content of the soil is expressed in mg (N) g–1 (soil) by: TN = 1.4 VA .
P

Controls Controls are made by distilling an ammonium sulphate solution of known concentration (1) one distillation after attack under the same conditions as the samples in the presence of 1,000°C fired quartz (blank assay), (2) one direct distillation of the ammonium sulphate solution. The value of the blank assay is calculated by the difference between the two. Two reference samples and two replicates of samples chosen randomly in the series should be analysed each day. Notes First the experimental standard X 31111 (1983) and then the international standard NF ISO 11261 (1995) were published for the determination of soil total nitrogen by distillation after Kjeldahl mineralization. Discolouration is not a sign of the end of mineralization, but an indicator of the end of the oxidation of the humified coloured organic matter signalling the end of the production of foam. The conversion of the N organic products into ammonium obeys other criteria such as temperature, catalyst, time, and nature of the N compounds. The addition of potassium sulphate enables the boiling point of the sulphuric acid to be increased and the attack time to be reduced. To avoid losses, use dose less than 0.5 g K2SO4 per millilitre of sulphuric acid with selenium as catalyst. The boiling point of H2SO4, normally 330°C, increases to 364°C after addition of 1 g of K2SO4 per mL of sulphuric acid. Catalysts (1) mercuric oxide is very effective but pollute the environment; mercuric oxide can form amino complexes which are not released during distillation unless the sample is not treated with thiosulfate, (2) selenium allows quantitative recovery of N; the temperature should not be too high (<367°C) and attacks should be limited in time (maximum 3 h) to avoid the risk of losses, (3) the NF ISO 11261 standard (1995) recommends replacing Se by anatase TiO2 which is less polluting for the environment (catalytic mixture 200 g K2SO4, 6 g CuSO4,5H2O, 6 g TiO2).

C, N (H, O, S)

349

Grinding the samples to 125 µm enables micro-methods to be used without a serious reduction in precision (Parnas and Wagner, 1921; Markham 1942), thereby reducing the cost price, pollution, and the extent of work surface required. The use of heating blocks decreases the turbulence of the attacks as heating is more regularly distributed. Programming the stages of heating makes it possible to accomplish the mineralization cycle without monitoring. 10.2.7 Kjeldahl N, Titration by Spectrocolorimetry Principle After digestion in a sulphuric acid medium catalysed by selenium (without copper or titanium) the reaction (Berthelot 1859; Bolleter et al. 1961) is continued in basic medium (sodium hydroxide buffer and dibasic sodium phosphate). The NH4+ ion reacts with the sodium hypochlorite and the sodium phenolate (or the sodium salicylate) to give a blue–green complex. The reaction is catalysed by the sodium nitroprussiate.
NH3 + ClONH2Cl
,
-

Equipment – cf. “Equipment” under Sect. 10.2.6 – Centrifuge – Automated segmented continuous-flow analysis with NH4+ manifold Reagents All the reagents should be of reference analytical grade: – Mineralization products: see “Reagents” under Sect. 10.2.6. – Brij 35 detergent: polyoxyethylene lauryl ether C12H25(OCH2CH2)OH (to limit contamination of the analytical manifold). – 20% sodium hydroxide (NaOH) stock solution. – Sodium phosphate, Na2HPO4,2H2O.

350

Organic Analysis

– 20% sodium potassium tartrate, NaKC4H4O6,4H2O (Seignette salt) stock solution. – Stock buffer solution: dissolve 89 g Na2HPO4,2H2O in 800 mL deionised water; cool and add 50 mL of 20% NaOH solution; homogenize and complete to 1,000 mL. – Buffer solution: mix 200 mL of buffer stock solution, 250 mL of stock solution and 20% of sodium potassium tartrate; add 60 mL of 20% NaOH stock solution and about 300 mL water; homogenize and cool; add 1 mL of Brij 35; bring to 1 L and homogenize. – Sodium salicylate: dissolve 150 g of sodium salicylate and 300 mg of sodium nitroprussiate in 800 mL water; bring to 1 L; filter on a Büchner funnel with a blue filter without NH4 and add 1 mL of Brij 35; store in a brown bottle protected from the light. – Sodium nitroprussiate, Na2Fe(CN)5NO,2H2O. – Sodium hypochlorite, NaClO: add 5 mL NaClO in 80 mL distilled water, bring to 100 mL; add 2 drops of Brij 35; prepare a fresh solution each day. – Ammonium sulphate, (NH4)2SO4. – Hydrochloric acid, HCl 37%. – Standard NH4: pour 80 mL HCl 37% in 800 mL water; cool and bring to 1,000 mL with water; prepare a stock solution of ammonium sulphate at 100 mg NH4+–N mL–1; store in the refrigerator; solutions should be freshly prepared each day starting from the stock solution to provide ranges from 0 to 50 µg mL–1. Procedure
-

Fig. 10.2. Titration of total nitrogen by segmented continuous-flow analysis after mineralization (Buurmans et al. 1996)

C, N (H, O, S)

351

Mineralization is carried out as in “Mineralization” in “Procedure (Macro-Method)” under Sect. 10.2.6, but with no copper sulphate in the catalyst mixture. At the end of the attack, cool the sample, and after complete cooling, dilute and pour into a 250 mL volumetric flask. After adjusting the volume and homogenisation, centrifuge each aliquot at 3,000 g for 15 min to obtain a clear solution. Remove the test samples for analysis from the supernatant and dilute if necessary in suitable stages of dilution (for example five times) when introducing the sample in the N–NH4 manifold. Titration is by absorption at 660 nm in a 15 mm colorimetric cell. The manifold proposed by Buurmans et al. (1996) is shown in Fig. 10.2 without the dilution stage. In the case of highly calcareous soils, it is sometimes necessary to add EDTA in addition to K and Na tartrate to avoid side effects due to the calcium which is precipitated in the sulphuric acid medium and can give a colloidal precipitate that is not easily visible to the naked eye (Gautheyrou and Gautheyrou 1965).

10.2.8 Kjeldahl N, Titration by a Selective Electrode
Principle The principle is similar to that described in Sect. 10.3.2 of Chap. 28 for the titration of ammonium nitrogen. Only the conditions of use are a little different. Here the gas diffusion electrode enables the results to be read directly on the calibration curve. The response is linear in the 0.5–500 mg L–1 zone of concentration of nitrogen with a lower degree of precision for high concentrations because of the logarithmic response. A range of concentrations of from 0.5 to 10 mg L–1 is recommended. This method has the advantage of being extremely simple and a wide range of concentrations can be analysed. Daily calibration is not necessary if the working temperature is always the same. Under normal working conditions, no interference occurs. Equipment – Mineralization equipment: cf. “Equipment” under Sect. 10.2.6. – Orion model 95–100 gas electrode. – Ionometer (pHmeter, mVmeter) with a resolution of 0.1 mV. – Thermostated bath regulated at 25°C. – Magnetic stirrer.

352

Organic Analysis

Reagents – Mineralization acid: cf. “Reagents” under Sect. 10.2.6. – Ten mole (NaOH) L–1 soda solution. – Standard solution: dissolve 1.179 g of ammonium sulphate ((NH4)2SO4, mw = 132.12) in distilled water and bring to 250 mL; this solution contains 1 mg (N) mL–1; dilute 10 times to obtain the initial reference solution 0.1 mg (N) mL–1. – Ammonium chloride solution to fill and maintain the electrode: 0.1 mol (NH4Cl) L–1. Procedure Mineralization is carried out as in “Mineralization” in “Procedure (Macro-Method)” under Sect. 10.2.6. At the end of the attack, cool the sample and transfer it to a 200 mL volumetric flask. After complete cooling, bring the volume to 200 mL. Construction of the standard curve: in 50 mL beakers add the volumes of the 0.1 mg (N) mL–1 reference solution listed in Table 10.4; complete to 20 mL with additional volumes of the blank attack solution (cf. Sect. 10.2.). Bring the temperature of the beakers and the blank to 25°C, immerse the electrode in the blank avoiding the presence of air under the membrane; add 2 mL of the 10 mol (NaOH) L–1 soda solution; start gentle agitation to limit the Vortex effect; after 5 min, adjust the zero of the ionometer.
Table 10.4. Range of titration of total nitrogen by ionometry; the content of the soil is given for a test sample of 2 g, volume of solution after attack (cf. “Mineralization” in “Procedure (Macro-Method)” under Sect. 10.2.6): 200 mL, aliquot for titration: 20 mL mL of standard solution 0.1 mg (N) mL–1
1 2 4 6 8 10

mL of attack mL NaOH solution for 10 mol L–1 20 mL
19 18 16 14 12 10 2 2 2 2 2 2

solution concentration mg (N) L–1
5 10 20 30 40 50

soil concentration mg (N) g–1
0.5 1 2 3 4 5

Proceed in the same way for the different points of the calibration range. Record the signal in mV and plot the calibration curve with the logarithms of nitrogen concentrations on the x-coordinate and mV on the y-coordinate.

C, N (H, O, S)

353

The measurements should be carried out on 20 mL aliquots of each attack solution using the same technique as for the calibrations. It is essential to carry out titrations immediately after the addition of soda to avoid loss of ammonia. Remarks Before starting analysis, it is important to check the pH of the measurement solutions which must be from 11 to 13 after the addition of soda (Table 10.4). In the case of a lower pH, slightly increase the volume of the 10 mol (NaOH) L–1 soda solution. After use, the electrode should be stored carefully following the manufacturer’s instructions. The filling solution should be renewed periodically. 10.2.9 Mechanization and Automation of the Kjeldahl Method The Kjeldahl method has the advantage of requiring only simple equipment and of being the reference method. However, manual methods are time consuming. They require large laboratory work surfaces and fume hoods to evacuate the heavy vapours. The technique can be improved by automation which makes it possible to avoid direct handling of dangerous reagents such as boiling sulphuric acid or concentrated soda. The use of programmable heating blocks (e.g. Technicon, Skalar, Tecator) enables the temperatures of mineralization to be regulated and sudden starts and foaming during the attacks to be limited. Partially automated equipment exists comprising management stations for mineralization, distillation, and titration at the macro and micro scale (e.g. Bicasa, Bucchi, Gerhart, Skalar, Foss-Tecator, Velp). Prolabo, Cem, Questron enable accelerated mineralization by microwave heating; the attack containers are processed automatically. Complete automation was provided by manufacturers such as Tecator, Foss, Perstop, or Gerhart. Mineralization is automated, and final titration is carried out by titrimetry with potentiometric detection of the end point of titration, or by spectrocolorimetry. The job of the analyst is limited to filling up the reagents, regulating the heating programmes and recovering the results at the end of the day. Monitoring is simplified because an alarm goes off in the case of accident.

354

Organic Analysis

10.2.10 Modified Procedures for NO3–, NO2–, and Fixed N These methods are seldom used in repetitive analyses because of the uncertainty of the results, the length of time required for the procedures, and the fact that the contents are often not very significant. Nitrate and Nitrite The determination of total N using wet methods does not take nitrate into account, but as it is generally not present in large quantities in natural environments, there will be no effect on results obtained on dried samples. However, in the case of cultivated soils dressed with manure with a high N content, it is preferable to rapidly check for the presence of nitrate using commercial tests, because nitrate and nitrite can introduce a random error. Nitrate and nitrite can be included in Total-N by means of redox sequences before Kjeldahl mineralization (1) by KMnO4/Fe2+ where the nitrite is oxidized into nitrate by a potassium permanganate solution, the nitrate then being reduced in ammonium by ferrous iron before the mineralization sequence in the traditional method (cf. “Procedure (Macro-Method)” under Sect. 10.2.6); (2) by a system using salicylic acid/ammonium thiosulfate in which nitration by salicylic acid can occur quantitatively only in absence of water (Fuson 1962); the nitrated compounds are then reduced by ammonium thiosulfate before Kjeldahl mineralization as in “Procedure (Macro-Method)” under Sect. 10.2.6 (Nelson and Sommers 1980; Du Preez and Bate 1989). The international NF ISO 11261 standard (1995) recommends (1) action of 4 mL of a salicylic/sulphuric acid mixture (25 g salicylic acid in 1 L concentrated H2SO4) for a few hours or overnight, (2) the addition of 0.5 g of sodium thiosulfate with moderate heating until the foam disappears, (3) cooling the flask and the addition of a catalyst (dioxide of titanium instead of selenium) and sulphuric acid for Kjeldahl mineralization. Nitrate and nitrite can also be titrated separately using the methods described in Chap. 28. Fixed or Occluded N The determination of fixed or occluded N is very delicate. Indeed, some authors have raised questions both with respect to its nature (inorganic NH4+–N only, or organic and inorganic N) and to its modes of fixing N. Fixed N is not titrated quantitatively by the traditional method particularly in soils containing 2:1 clays. To titrate fixed N, a polyethylene bottle is used, and the clay lattice is destroyed by a mixture HF-KCl or concentrated KOH (cf. Sect. 28.3.5 of Chap. 28). The organic digestion of N is then

C, N (H, O, S)

355

carried out in traditional Kjeldahl flasks (cf. Sect. 10.2.6).
Remarks As the glass of the Kjeldahl flasks is seriously affected, their quality deteriorates rapidly and they need to be regularly replaced. The fixing of N–NH4 in the mineral lattices is more important in the deep horizons which are rich in clay and low in N compounds; the increase in N-fixing power as a function of the nature of clays can be expressed by: vermiculites > illites > montmorillonites > kaolinites. In certain cases the expansion of the lattices, which is obtained by the addition of water, can lead to higher values for total-N (Bal 1925). Conversely, the rate of recovery of NO3_N may be lower (Bremner and Mulvaney 1982).

10.3 Dry Methods
10.3.1 Total Carbon by Simple Volatilization These methods are based on oxidation of the soil organic matter and on destruction of carbonates by heating of the samples to relatively high temperatures for a given length of time. Measurement of the loss in mass (cf. Chap. 1) allows estimation of gaseous losses that are mainly in the form of CO2 and H2O. Resistance or induction furnaces can function at temperatures of 1,000–1,500°C. In the simplest case, an open air circuit is used, possibly with oxygen enrichment (low temperature ashing, LTA), or in closed systems with controlled temperature and time. Traps can be used to block some compounds, and catalysts can accelerate the reactions. Gas separation systems of varying degrees of sophistication are used in sequential analyses of evolved gas. To sum up, a temperature range lower than 500°C enables collection of the CO2 released by the organic matter and temperatures higher than 500°C, of CO2 produced from carbonates. Although calcination methods using an open circuit for the determination of Total-C are very simple to implement, they cannot be regarded as quantitative and generally have be supplemented by thermal analysis techniques (cf. Chap. 7). In an oxidizing atmosphere, gravimetric measurements will represent: – Organic C transformed into CO2. – Inorganic C starting from approximately 400°C.

356

Organic Analysis

– Different forms of water (hygroscopic, interstitial, water of crystallization, hydroxyl groups) which are moved throughout the thermal cycle. – Volatile forms of N, S, certain metals, and metalloids (Cl–). These displacements vary with the temperature used and with the times of application. Still other transformations can disturb measurements. In an oxidizing atmosphere, some compounds of the soil that are sensitive to redox reactions will increase in weight while passing to a higher valence, for example Fe2+ → Fe3+. Sulphur compounds can, by recombination, show simultaneous losses and increases in mass as a function of the medium: H2S → SO2 → SO3 → SO42–. The results are thus random, and are related to the atmosphere used (oxidizing, neutral, reducing) and to effects of the matrix. The thermostable residue of the soil can be considered reached between 1,100 and 1,600°C. In most cases, total organic and inorganic C with C obtained by the simple measurement of the losses on ignition cannot be directly assimilated. Below 500°C, C losses at lower temperatures can easily characterize oxidizing C, but it is difficult to quantify this variable. However, certain soils enable results with a reduced margin of error, in particular sandy soils with low 1:1 clay content, but not calcareous soils, or soils rich in organic matter. The decarboxylation of the organic matter starts around 180°C, and at 250°C, it is estimated that 70% of organic C is oxidized, and 90% around 500°C. Temperatures between 800 and 950°C are needed to obtain 99% of oxidation of organic C. In the case of carbonates, decomposition starts at around 400°C: calcium and magnesium carbonates (e.g. calcite-aragonite, magnesite, dolomites) generally display the same type of thermal behaviour, but biogenic calcite (shells, skeletons, calcareous debris), iron carbonate (siderite), fossil coals, and resins can respond differently, as can sodium carbonates with a low melting point (851°C). 10.3.2 Simultaneous Instrumental Analysis by Dry Combustion: CHN(OS) Principle Thanks to improved equipment, the purchase of entirely automated CHN(OS) apparatuses (Pansu et al. 2001) is very tempting. These apparatuses are very profitable in laboratories that do many repetitive

C, N (H, O, S)

357

analyses, as they can do simultaneous analysis of C, N, and H and, under adequate operating conditions, of O and S in addition. To ensure the uninterrupted performance of the apparatuses, the operators need to be specially trained. The apparatuses also have to be regularly checked and maintained, and the exact conditions of use clearly defined. In this type of set-up, each element in the chain is linked to the others (furnace, CuO, catalysts, and different traps) and the weakest link in the chain determines the final quality of the analyses. The method is the subject of the international standard NF ISO 10694 (1995) for the analysis of organic carbon and total carbon of soils. Seal a given mass representative of the sample in a tin (or silver) micro-capsule and place it in a furnace at 1,100°C in a controlled atmosphere in the presence of oxygen and copper oxide. The different forms of organic and inorganic C and N oxidize or break down rapidly according to the principle of the Liebig reaction (1830)
O2 ,1, 800° C C + 2CuO ⎯⎯⎯⎯→ CO2↑ + 2Cu

The temperature of 1,800°C is reached by flash combustion of the tin capsules. The carbonates and bicarbonates are broken up simultaneously. For titration of organic C, carbonates, and bicarbonates must be eliminated chemically before the sample is placed in the furnace
1,000° C CaCO3 ⎯⎯⎯→ CO2↑ + CaO CaCO3 + 2H+ → CO2↑ + H2O + Ca2+

(thermal destruction) (chemical destruction)

If methane is present and has to be titrated, Cr2O3 should be used to ensure perfect oxidation. The organic nitrogen that is oxidized during combustion gives nitrogen oxides which must be reduced with copper to transform all nitrogen oxides into the N2 form by the Dumas (1831) method. Inorganic N compounds (NH4+, NO3_, NO2_) are subtracted to obtain total organic N. Different catalysts can be used to accelerate the reactions or to make them quantitative. Finally, the gas products are identified in suitable detectors; the gaseous compounds can be separated on temporary specific traps or by gas chromatography, depending on the system used. In soils that do not contain carbonates or bicarbonates, or have not received calcareous amendments or lime: – Inorganic C = 0. – Total C = total organic C.

358

Organic Analysis

Equipment – Elementary Analyzer CHN(OS), preferably able to handle samples of approximately 100 mg or more to compensate for the heterogeneity of carbon in soil (Pansu et al. 2001)1. – Micro analytical balance. – Needle-nosed pliers to close the capsules. – Micro pipette adjustable 0.1–0.5 mL. – Ceramic plate for treatment of calcareous samples. – Controlled temperature hotplate. Products All consumables should be suitable for the procedure and type of analytical materials used (analytical grade reference products): – Combustive gases, carrier gas purity: 99.98% – Copper oxide, (CuO) – Copper in the form of wire or chips (Cu) – Magnesium perchlorate (Mg(ClO4)2, ascarite (NaOH-asbestos), MnO2, etc. for traps – Tin or silver micro capsules, capacity 100 mg – A range of catalysts – Standard substances for calibration and control such as acetanilide (C8H9ON, 71.09% C, 6.71% H, 10.36% N, 11.84% O), atropine (C17H25NO3), sulphanilamide H2NC6H4SO2NH2, picric acid (2,4,6 trinitrophenol (NO2)3C6H2OH), thiourea (H2NCSNH2), etc. Procedure Non-calcareous soils All weighing should be done on analytical balances (± 0.1 or ± 0.01 mg) Calibration: weigh 5 sample specimens of increasing mass of a standard in the range of concentration accepted by the CHN apparatus concerned. The precision of the equipment should be tested by comparing the theoretical contents with the contents obtained.

1

A higher mass of sample specimen increases the representativeness of the sample but produces more ashes in the combustion furnace. The current trend is to reduce the mass of sample specimen to about 10 mg and to increase the homogeneity of the sample by very fine grinding to a particle-size of less than 100 micrometers.

C, N (H, O, S)

359

Samples: grind the samples to 125 µm (Sieve AFNOR NF22) and dry for 48 h in a P2O5 desiccator to limit saturation of the traps. In tin or silver capsules, carefully weigh from 50 to 100 mg of sample (depending on the estimated contents of C and N) dried on P2O5, and seal the capsule. Place the capsules in the sample distributor of the CHN apparatus. At the same time, weigh 1 g of the same sample to measure residual moisture by drying at 105°C in order to be able to correct the results (cf. Chap. 1). Direct drying of the samples at 105°C before CHN analysis may increase ammonium fixation in the lattice of 2:1 clays, but in most case this is acceptable. Note Andosols and histisols pose two problems (1) the residual moisture of airdried soil is very high, up to 50–60% after 6 months; (2) low bulk density can reach 0.30–0.25. In this case it is not possible to weigh 100 mg in the capsules. Carbon in calcareous soils In the case of calcareous samples or soils that have been limed, if there is no siderite (not easily decomposable FeCO3), treatment with 10% hydrochloric acid solution is sufficient to destroy the carbonates and bicarbonates in the sample: – Weigh 100 mg of sample in a silver capsule. – Place the open capsule on a ceramic plate. – With a micro pipette, slowly add 0.1–0.2 mL of 10% HCl depending on the estimated quantities of carbonate and bicarbonate. – Leave in contact for 2 h then add 0.1 mL of the 10% HCl solution. Check that gaseous emission is complete. – Place the ceramic plate on a hotplate at 60°C and bring to dry. Leave to cool and seal the capsule. Place the capsule on the CHN sampler to measure total organic C. A second sample can be weighed without destroying carbonates to measure total C. Total inorganic C can be calculated by difference. On certain apparatuses it is preferable to use phosphoric acid rather than hydrochloric acid to limit the effects of chloride on the catalysts and on the traps.

360

Organic Analysis

Calculations The results are calculated directly and printed by the CHN analyser after measurement of the surface of the C and N peaks. These calculations take the weight of the sample into account. To determine total organic nitrogen, the results obtained for total inorganic nitrogen have to be taken into account (cf. Chap. 28). Total N (CHN) = inorganic Nitrogen + organic Nitrogen Titration of Hydrogen, Oxygen and Sulphur Hydrogen is titrated together with carbon and nitrogen using the procedure described in “Procedure” under Sect. 10.3.2 by measuring the water formed during combustion. However this measurement is difficult to interpret in most soils because hydrogen resulting from combustion of the organic molecules and hydrogen coming from the different forms of water (e.g. adsorption, hydration, and hydroxylation) is not separated in the H2O signal. In soils that are not very clayey and when using carefully dried samples, the H2O signal can represent only H coming from combustion of the organic molecules. The C:H ratios then reveal the stages of evolution of the soil organic matter. The titration of oxygen and sulphur is generally carried out with standard CHN equipment by simply modifying the instrumental parameters. For sulphur, controlled oxidation is used, generally in the presence of tungstic anhydride. For oxygen, pyrolysis is used instead of combustion (the oxygen supply is cut off) on a sample specimen reserved for this measurement. Total organic oxygen: the sample is subjected to pyrolysis at 1,100°C in the presence of a catalyst such as a mixture of carbon and nickel. Carbon monoxide (CO) is formed, then isolated by a separation system (generally with a trap or by chromatography) and quantified by a system of detection (e.g. a catharometer). Oxygen is determined by comparison with standards of known O content in the same way as for C and N described in “Procedure” under Sect. 10.3.2: one molecule of CO represents one O atom in the sample. Total organic oxygen corresponds to only a small fraction of total oxygen of the soil. Inorganic oxygen, which is more abundant, is generally estimated starting from the total analysis of the major elements (cf. Chap. 31), by calculating the difference between the sum of the contents expressed in oxides and the sum of the contents expressed in elements. Total sulphur: the sample is oxidized at 1,100°C in the presence of tungstic anhydride WO3 or a mixture of tungstic and vanadic oxides.

C, N (H, O, S)

361

Oxides of sulphur (e.g. sulphur trioxide SO3, sulphur dioxide SO2) are formed. A stage of reduction on a copper column transforms all oxides into the SO2 form. This gas is then isolated and titrated using a system that will depend on the type of equipment used (traps, chromatography) and the type of detector (catharometer, IR detector): one molecule of SO2 corresponds to one atom of S in the sample. Depending on the type of equipment used, titration of sulphur may or may not be carried out simultaneously with that of carbon and nitrogen. Instrumental CHNS methods give a value for the sulphur content that can generally be regarded as the total sulphur content of the soil: oxidation of organic sulphur and sulphides, decomposition of most of the sulphates (Laurent 1990). However, care should be taken given the diversity of natural forms of sulphur. For a more detailed analysis of this element, see Chap. 30. Remarks Alkaline salts (e.g. sodium carbonate, chlorides, phosphates) melt to a varying degree and delay the oxidation of C into CO2 by coating the organic particles and disturbing the cycle of analysis. Salts that are volatile at 700°C or more (e.g. some chlorides, bromides, or iodides) can contaminate (or corrode) the circuits, traps, and catalysts. Magnesium perchlorate Mg(ClO)4)2, sold under the name of anhydrone or dehydrite, is used to trap water. It is a relatively unstable product that, after use, must be eliminated in the same way as other dangerous waste by the laboratory. Residues of calcination in the furnace have to be removed frequently, particularly in the case of carbonated saline soils. Changing the protective nacelles makes it possible (1) to preserve the quality of the filters for elimination of the fine particles likely to be present in the gas phase and (2) to limit the retention of volatile products of combustion. Traps, compounds, catalysts, and filters must be changed regularly, respecting the change-by deadlines. Most breakdowns and doubtful results are due to inadequate maintenance. The use of synthetic standards can result in errors, as their physical and chemical behaviour may differ from that of soil organic matter. However, the temperature of 1,800°C reached during the flash combustion of the capsules makes it possible to release all the organic compounds. Soil standards with certified organic-C contents are also available for calibration. Elementary CHN(OS) analysers generally give accurate results but these may be higher than wet oxidation.

362

Organic Analysis

10.3.3 CHNOS by Thermal Analysis In detailed studies it is important not to overlook the possibilities offered by instrumental methods such as differential thermal analysis (DTA) and thermogravimetric analysis (TGA), coupled or not with measurements of evolved gas analysis (EGA) (cf. Chap. 7).

Water losses from gypsum allophane

Oxidation of nucleus of humic acids Water losses from clays, oxidation of functional groups Biogenic calcite of humic acids

Calcite

Magnesite

Aragonite

0

100

300 Mostly oxidation of organic C

500

700

900

T°C

Mostly decomposition of inorganic C

Fig. 10.3. Oxidation of organic C and decomposition of inorganic C as a function of temperature

Controlled pyrolysis (with suitable temperature/time programmes) enables oxidation of the organic matter and the decomposition of the inorganic compounds to be monitored. It is possible to identify and quantify the nature of the gases (e.g. CO2, CO, CH4, H2O, NH3, H2S, SO2) that correspond to the DTA and TGA peaks as a function of temperature, to monitor the decomposition or the transformation of products containing C and N, and to separate exothermic and endothermic peaks of H2O and CO2, etc. For more complete studies, trapping of certain phases at –180°C, then their separation by gas chromatography, or by coupling with, for

C, N (H, O, S)

363

example, infra-red or mass spectrometers detectors, enables detailed characterization of soil organic matter. The sum of the different carbon phases gives total C; and it is possible to roughly separate total organic C and total inorganic C (Fig. 10.3). In studies of organic geochemistry, the molecular mass of the fragments of pyrolysis is measured in a mass spectrometer with discontinuous injections. The structure of complex substances with high molecular weights can be characterized.

10.3.4 C and N Non-Destructive Instrumental Analysis
Analysis by diffuse reflectance near infra-red spectrometry (NIRS, cf. Chap. 5) is carried out directly on the soil sieved on an AFNOR NF22 sieve (125 µm mesh). Take a sample weighing approximately 1 g after drying in the drying oven and carefully pack it in a special cup and place the cup in the measuring chamber. The surface of the sample must be perfectly flat. Each component of the organic complexes of the soil has a specific absorption point between 700 and 2,500 nm in the NIR spectrum, due to the vibrations of stretching and deformation of the inter-elements bonds (cf. Sect. 10.3.1 in Chap. 5). For C and N measurement, information provided by the NIRS spectra must be calibrated to data from other methods of measurement. Then the calibration curve is used to quantify C and N in unknown samples. Good calibration has been obtained with the methods described in Sects. 10.3.1–10.3.3 above (Krishnan et al. 1980; Dalal and Henry 1986; Morra et al. 1991; Fidêncio et al. 2002). The method is non-destructive and the samples can be used for other measurements. As each apparatus has its own particular characteristics with regard to selection and optimization, the manufacturer’s recommendations should be followed (e.g. BranLuebbe, Bruker, Foss, Leco, Nicolet, Perkin-Elmer, and Perstorp). Rapid methods to estimate soil C in the field are currently the object of serious investigation in research programmes dedicated to C sequestration in soils as part of the effort to decrease emissions of greenhouse gases in the atmosphere. Though they are less precise than laboratory techniques using a wet method or dry combustion, these methods have the advantage of rapidly providing a very large number of measurements at a lower cost. NIRS methods for the processing of analytical signals have been developed thanks to spectacular progress in software. This software can provides quantitative data based on spectra that chemists previously found very difficult to interpret.

364

Organic Analysis

In addition to the NIRS method, the laser induced breakdown spectroscopy method (LIBS) has been proposed for soil carbon (Cremers et al. 1996). The detection limit is approximately 300 mg kg–1, precision 4–5%, accuracy 3–14%, and the speed of analysis is more than one sample per minute (Cremers et al. 2001).

10.3.5 Simultaneous Analysis of the Different C and N Isotopes

Fig. 10.4. Simultaneous determination of total- and 14C-carbon (diagram by P. Bottner, CEFE-CNRS Montpellier, France, personal communication), (a) soda lime for O2 purification, (b) boiling liquid sample + K2Cr2O7, (c) combustion tube, (d) post combustion tube (CuO), (e) solid sample, (f) 900°C combustion furnace, (g) automated burette containing H2SO4 + H3PO4, (h) purification traps (water condensation, chromic and acid foams), (i) columns with glass balls impregnated with ethylene glycol mono-ethyl ether + mono-ethanolamine to trap CO2, ( j) soda lime traps for security, (k) ventilated external output, (l) output of excess O2

Dry combustion analysers – CHN(OS) – can be coupled with mass spectrometers enabling the study of the different isotopes of the elements, particularly 14C, 13C, 15N (Pansu et al. 2001). Often 15N nitrogen is also measured starting from the Kjeldahl distillates in the form of ammonium. The carbon dioxide of the effluent combustion gas can also be trapped for later determination of the isotopes. The Bottner and Warembourg

C, N (H, O, S)

365

equipment (1976), shown in schematic form in Fig. 10.4, was used for more than 20 years for simultaneous titration of total-C (carmhograph 12A) and 14C-carbon (liquid scintillation) starting from solid (soils, plants) or liquid (water of the soil) samples and proved to be reliable.

References
Anne P (1945) Sur le dosage du carbone organique des sols Ann. Agron., 15, 161–172. Baker KF (1976) The determination of organic carbon in soil using a probe colorimeter. Lab. Practice, 25, 82–83. Bal DV (1925) The determination of nitrogen in heavy clay soils. J. Agric. Sci., 15, 454–459. Berthelot MP (1859) Violet d’amiline. Répertoire chim. Appl., 1, 284. Bolleter WT, Bushman, CJ and Tidwell PW (1961) Spectrophotometric determination of ammonia as indophénol. Anal. Chem., 33, 542–594. Bottner P and Warembourg F (1976) Method for simultaneous measurement of total and radioactive carbon in soils, soil extracts and plant materials, Plant Soil, 45, 273–277. Bremner JM and Mulvaney CS (1982) Nitrogen total. In Methods of soil analysis- part 2. Chemical and microbiological properties. ASA-SSSA, 9, 595–624. Buurmans P, van Lager B and Velthorst FJ (1996) Manual for soil and water analysis., Backhuis Publishers, Leiden, 17–21. Cremers DA, Ferris MJ and Davies (1996) Transportable Induced Laser Breakdown Spectroscopy (LIBS) instrument for field-based soil analysis. Proc. Soc. Photo Opt. Inst. Eng., 2835, 190–200. Cremers DA, Ebinger MH, Breashears DD, Uukefer PJ, Kammerdiener SA, Ferris MJ, Catlett KM and Brown JR (2001) Measuring Total Soil Carbon with Laser-Induced Breakdown Spectroscopy (LIBS). J. Env. Qual., 30, 2202–2206. Dalal RC and Henry RJ (1986) Simultaneons determination of moisture, organic carbon and total nitrogen by near infrared reflectance spectrometry. Soil Sci. Soc. Am. J., 50, 120–123. Du Preez DR and Bate GC (1989) A simple method for the quantitative recovery of nitrate N during Kjeldahl analysis of dry soil and plant samples. Commun. Soil Sci. Plant Anal., 20, 345–357. Fidêncio PH, Poppi RJ and de Andrade JC (2002) Determination of Organic Matter in Soils using function networks and near infrared spectroscopy. Anal. Chim. Acta., 453, 125–134. Fuson RC (1962) Reactions of organic compounds - A text book for the advanced student. Wiley, Bristol 1962.

366

Organic Analysis

Gautheyrou J and Gautheyrou M (1965) Dosage simultané de l’azote ammoniacal et nitrique dans le sols. Contribution à l’étude de la dynamique de l’azote. Cah. Orstom Sér.Pédol., III, 367–391. Keerrthisinghe GK, Mengel K and De Datta SK (1984) The release of nonexchangeable ammonium (15N Labelled ) in wetland rice soils. Soil Sci. Soc. Am. J., 48, 291–294. Kjeldahl J (1883) Neue méthode zur bestimmung des stickstoffs in organischen körpern. Z. Anal. Chem., 22, 366–382. Krishnan P, Alexander JD, Butler BJ and Hummel JW (1980) Reflectance techniques for predicting soil organic matter. Soil Sci. Soc. Am. J., 44, 1282–1285. Laurent JY (1990) Analyse élémentaire d’échantillons de sol et solutions: application de la microanalyse au dosage simultané de l’azote, du carbone et du soufre. In Actes Journées laboratoire, Orstom, Bondy, France. Ma TS and Zuazaga G (1942) Micro - Kjeldahl determination of nitrogen. A new indicator and an improved rapid method. Ind. Eng. Chem. Anal., 14, 280–282. Markham R (1942) A steam distillation apparatus suitable for micro - Kjeldahl analysis. Biochem. J., 36, 790–791. Mebius LJ (1960) A rapid method for the determination of organic carbon in soil. Anal. Chem. Acta., 22, 120–124. Mengel K and Scherer HW (1981) Release of non exchangeable (fixed ) soil ammonium under field condition during the growing season. Soil Sci., 131, 226–232. Morra MJ, Hall MH and Free Born LL (1991) Carbon and nitrogen analysis of soil fraction using near infrared reflectance spectroscopy. Soil. Sci. Soc. Am. J., 55, 281–291. NF X31-109 (1993) Détermination du carbone organique par oxydation sulfochromique. In Qualité des sols, 1996, AFNOR, Paris, 67–73. NF ISO 10694 (1995) Dosage du carbone organique et du carbone total après combustion sèche (analyse élémentaire). In Qualité des sols, 1996, AFNOR, Paris, 189–199. NF ISO 11261 (1995) Dosage de l’azote total - Méthode de Kjeldahl modifiée. In Qualité des sols, 1996, AFNOR, Paris, 257–260. X 31-111 (1983) Détermination de l’azote total - Méthode par distillation après minéralisation ( Kjeldahl ). In Qualité des sols, 1994, AFNOR, Paris, 69–72. Pansu M, Sallih Z and Bottner P (1998) A process-based model for carbon and nitrogen transfers in soil organic matter. In Actes 16e congrès mondial de science du sol, 20–26 april, Montpellier, France Pansu M, Gautheyrou J and Loyer JY (2001) Soil Analysis - Sampling, Instrumentation and Quality control. Balkema, Lisse, Abington, Exton, Tokyo, 489. Parnas JK and Wagner R (1921) Ûber dieausfuhrung von bestmmungen kleiner stickstoffmengen nach Kjeldahl. Biochem. Z., 125, 253–256.

C, N (H, O, S)

367

Schollenberger CJ (1927) A rapid approximate method for determining soil organic matter. Soil Sci., 24, 65–68. Walkley A and Black A (1934) An examination of the Degtjareff method for determining soil organic matter and a proposed modification of the chromic acid titration method. Soil Sci., 37, 29–38. Winkler LW (1913) Beitrag zur titrimetrishen bestimmung des ammoniaks. Z Angew. Chem., 26, 231–232.

Bibliography
Allison LE, Bollen WB and Moodie CD (1965) Total carbon. In Black C. A. et A.L. Method of soil an Analysis part 2. Agronomy Monograph n°9. Am. Soc. Agron., 1346–1366. Alves BJR, Boddey RM and Urquiaga SS (1993) A rapid and sensitive flow injection technique for the analysis of ammonium in soil extracts. Commun. Soil Sci. Plant Anal., 24, 277–284. Association of Official Analytical Chemists. AOAC (1975) Official methods of analysis, 924–925. Bernoux M and Cerri CEP (2005) Soil organic components. In Encyclopedia of Analytical Science, 2nd Edition, Vol. 4, Elsevier, Amsterdam. Bowman RA, Guenzin WD and Savory DJ (1991) Spectroscopic method for estimation of soil organic carbon. Soil Sci. Soc. Am. J., 55, 563–566. Bremner JM and Tabatabai M.A (1972) Use of an ammonia electrode for determination of ammonium in Kjeldahl analysis of soils. Commun. Soil Sci. Plant Anal., 3, 159–165. Burton DL, Gower DA, Rutherford PM and McGill WB (1989) Amino acid interference with ammonium determination in soil extracts using the automated indophenol method. Commun. In Soil Sci. Plant Anal., 20, 555–565. Carr CE (1973) Gravimetric determination of soil carbon using the Leco induction furnace. J. Sci. Food Agric., 24, 1091–1095. Cheng HH and Kimble J (2000) Methods of Analysis for Soil Carbon: An Overview. In Global Climate Change and Tropical Ecosystems., CRC, Boca Raton, 333–339. Chichester FW and Chaison RF (1992) Analysis of carbon in calcareous soil using a two temperature dry combustion infra-red instrumental procedure. Soil Sci., 153, 237–241. De Bolt DC (1974) A high sample volume procedure for the colorimetric determination of soil organic matter. Commun. Soil Sci. Plant Anal., 5, 131–137. Decreider and Meaux R (1973) Utilisation d’une electrode ionique spécifique pour le dosage de l’azote par la méthode Kjeldahl. Analusis, 2, 442–445. Diaz-Zorita M (1999) Soil organic carbon recovery by the Walkley-Black method in a typic Hapludoll. Commun. Soil Sci. Plant Anal., 30, 739– 745.

368

Organic Analysis

Donkin MJ (1991) Loss - on - ignition as an estimator of soil organic carbon in A horizon forestry soils. Commun. Soil Sci. Plant Anal., 22, 233–241. Froelich PN (1980) Analysis of organic carbon in marine sediments. Limnol. Oceanogr., 25, 564–572. Gallaher RN, Weldon CO and Bowell FC (1976) A semi-automated procedure for total nitrogen in plant and soil samplers. Soil. Sci. Soc. Am. J., 40, 887–889. Genty CE and Willis RB (1988) Improved method for automated determination of ammonium in soil extracts. Commun. In Soil Sci. Plant Anal., 19, 721–737. Graham ER (1948) Determination of soil organic matter by means of a photoelectric colorimeter. Soil Sci., 65, 181–183. Guillet B (1979) Etude du renouvellement des matières organiques des sols par les radio isotopes 14C. In Pedologie - constituants et propriétés du sol, Bonneau M and Souchier B ed. Masson, Paris, 210–226. Heanes DL (1984) Determination of total organic C in soils by an improved chromic acid digestion and spectrophotometric procedure. Commun. Soil Sci. Plant Anal., 15, 1191–1213. Henzell EF, Wallis I and Lindquist JE (1968) Automatic colorimetric methods for the determination of nitrogen in digests and extracts of soils. Int. Cong. Soil Sci. Trans., 9th (Adelaïde ), 3, 513–520. Howard PJ A and Howard DM (1990) Use of organic carbon and loss on ignition to estimate soil organic matter in different soil types and horizons. Biol. Fert. Soils., 9, 306–310. Kalembasa SI and Jenkinson DS (1973) A comparative study of titrimetric and gravimeric methods for determination of organic carbon in soils. J. Sci. Food Agric., 24, 1085–1090. Kalisz PJ and Sainju UM (1991) Determination of carbon in coal “ blooms ”. Commun. Soil Sci. Plant Anal., 22, 393–398. Kempers AJ and Kok CJ (1989) Re-examination of the determination of ammonium as the indophenol blue complex using salicylate. Anal. Chim. Acta., 221, 147–155. Lal R, Kimble JM, Follet RF and Stewart BA, (2001) Assessment Methods for Soil Carbon, Lewis CRC, London/Boca Raton, 676. Lowther JR, Smethurst PJ, Carlyle JC and Nambiar EKS (1990) Methods for determining organic carbon in podzolic sands. Commun. Soil Sci. Plant Anal., 21, 457–470. Mc Geehan SL and Naylor DV (1988) Automated instrumental analysis of carbon and nitrogen in plant and soil samples. Commun. Soil Sci. Plant Anal., 19, 493–505. Merry RH and Spouncer LR (1988) The measurement of carbon in soils using a micro processor - controlled resistance furnace. Commun. Soil Sci. Plant Anal., 19, 707–720. Nelson DW and Sommers LE (1975) A rapid and accurate procedure for estimation of organic carbon in soil. Proc. Indiana Acad. Sci., 84, 456– 462.

C, N (H, O, S)

369

Nelson DW and Sommers LE (1982) Total carbon, organic carbon and organic matter. In Methods of soil analysis part 2. Page AL, Miller RH and Keeney DR ed. Agron. Monogr., 9, ASA - SSSA Madison., 539–579. Nelson DW and Sommers LE (1996) Total carbon, organic carbon and organic matter. In Methods of soil analysis, part 3, chemical methods, Sparks DL ed. SSSA Book series no 5, 961–1010. Nelson DW (1983) Determination of ammonium in KCl extracts of soils by the salicylate method. Commun. Soil Sci. Plant Anal., 14, 1051–1062. Nommik H (1971) A modified procedure for determination of organic carbon in soils by wet combustion. Soil Sci., 111, 330–336. Patton JC and Crouch SR (1977) Spectrophotometric and kinetics investigation of the Berthelot reaction for the determination of ammonia. Anal. Chem., 3, 464–469. Pella E (1990) Elemental organic analysis. Part 1, historical development, 116– 125; part 2, state of the art, 28–32, Ann. Lab. Perrier ER and Kellog M (1960) Colorimetric determination of soil organic matter. Soil Sci., 90, 104–106. Puttanna K and Prakasa Rao EVS (1993) Determination of nitrate in soil by second derivative ultraviolet spectrometry. Commun. Soil Sci. Plant Anal., 24, 737–743. Qiu Xing-Chu and Zhu Ying-Quan (1987) Sensitive spectrophotometric determination of ammoniacal nitrogen analysis. Analusis, 15, 254–258. Quinn JG and Salomon M (1964) Chloride interference in the dichromate oxidation of soil hydrolyzates. Soil Sci. Soc. Am. Proc., 28, 456. Rowland AP (1983) An automated method for the determination of ammoniumN in ecological materials. Commun. Soil Sci. Plant Anal., 14, 49–63. Schepers JS, Francis DD and Thompson MT (1989) Simultaneous determination of total C, total N, and 15N in soil and plant materiel. Commun. Soil Sci. Plant Anal., 20, 949–959. Schulte EE, Kaufman C and Peter JB (1991) The influence of sample size and heating time on soil weight loss - on - ignition. Commun. Soil Sci. Plant Anal., 22, 159–168. Schuman GE, Stanley MA and Knudsen D (1973) Automated total nitrogen analysis of soil and plant samples. Soil Sci. Soc. Am. Proc., 37, 480– 481. Sheldrick BH (1986) Test of the Leco CHN 600 determination for soil carbon and nitrogen analysis. Canadian J. of Soil Sci., 66, 543–545. Sims JJ and Haby VA (1971) Simplified colorimetric determination of soil organic matter Soil Sci., 112, 137–141. Skjemstad JO, Reeve R (1976) The determination of nitrogen in soils by rapid high temperature Kjeldahl digestion and auto analysis. Commun. Soil. Sci. Plant Anal., 7, 229–239. Soon YK and Aboud S (1991) A comparison of some methods for soil organic carbon determination. Commun. Soil Sci. Plant Anal., 22, 943–954. Stewart RA and Porter LK (1963) Inability of Kjeldahl methods to fully measure indogeneous fixed ammonium in some soil. Soil Sci. Soc. Am. Proc., 27, 41–43.

370

Organic Analysis

Szekely E (1991) A rapid colorimetric method for analysis of nitrate nitrogen by reduction to nitrite. Commun. Soil Sci. Plant Anal., 22, 1295–1302. Tabatabai MA and Bremner JM (1991) Automated instruments for determination of total carbon, nitrogen and sulphur in soils by combustion techniques. In Soil analysis - Modern instrumental techniques, Smith KA ed. Marcel Dekker, NY, USA 261–286. Takesako H (1991) Double-plunger pump system flow injection spectrophotometric determination of inorganic nitrogen in soil extracts, part 1, flow injection analysis of ammonium nitrogen in soil extracts, part 2, flow injection analysis of nitrate nitrogen in soil extracts, Jpn J Soil Sci. Plant Nut., 62, 128–134, 135–140 Tel DA and Jansen J (1992) Determination of total nitrogen in soil digest using a Traacs 800 Autoanalyser. Commun. Soil Sci. Plant Anal., 23, 2729– 2736. Tisley J (1950) Determination of organic carbon in soils by dichromate mixtures in Trans. 4 th Int. Cong. Soil. Sci., 1, 161–169. Hoitsemo Brothers, Groningen. Verardo DJ, Froelich PN and Mcintyre A (1990) Determination of organic carbon and nitrogen in marine sediments using the Carlo Erba Na. 1500 analyser. Deep Sea Res., 37, 157–165. Volonteri HJ (1983) Modificacion de un metodo colorimetrico para la determinacion del nitrogeno total en suelos y pastos. Ciencia del suelo, 1, 98–99. Walinga J, Kithome M, Novozamsky I, Houba VJG and Vanderlee JJ (1992) Spectrometric determination of organic carbon in soils. Commun. Soil Sci. Plant Anal., 23, 1935–1944. Winter JP, Gregorich EG, Voroney RP and Kachanoski RG (1990) Comparison of two samples oxidation methods for quantitative measurement of 12C and 14C in plant and soil. Canad. J. Soil Sci., 70, 525–529. Yeumans JC and Bremner JM (1991) Carbon and nitrogen analysis of soils by automated combustion techniques. Commun. Soil Sci. Plant Anal., 22, 843–850.

11 Quantification of Humic Compounds

11.1 Humus in Soils

11.1.1 Definition Humus plays a fundamental role in ecological processes as a source of carbon for the atmosphere, a sink of carbon for the biosphere, a sink and source of fertilizers for plants, and a factor that influences soil properties, and important reviews are regularly devoted to it (Kononova 1966; Flaig et al. 1975; Schnitzer 1978; Stevenson 1982; 1994; Aiken et al. 1985; Tate 1992; Carter and Stewart 1995; Piccolo 1996; Magdoff et al. 1996; Hessen and Tranvik 1998). Stevenson (1982) defined the term “humus” (or humified matter) as the sum of organic compounds in the soil with the exclusion of living organisms in the biomass and non-decomposed or partially decomposed organic debris of plant or animal origin (cf. Chap. 9). The use of the term “soil organic matter” is less clear: the term is sometimes used with the same meaning as humus but in fact should refer to the sum of soil organic materials. Humified matter represents more than half soil total organic carbon and can be classified in two main types, humic and non-humic substances (Schnitzer 1978). The physical and chemical characteristics of non-humic substances e.g. carbohydrates, proteins, peptides, amino acids, lipids, waxes and organic acids of low molecular weight are easily recognizable. Humic substances on the other hand, do not show such marked physicochemical characteristics. They are more or less dark in colour, their molecular weight varies from a few hundred to several hundred thousand daltons, and they display a complex chemical structure, a hydrophilic character and acid properties.

372

Organic Analysis

However, the distinction between the two types of substances is not completely clear since humic substances always contain non-humic substances, which can be released by chemical treatments like acid hydrolysis.

cf. Chap.9

2,000 d 1,400 Cmol(H+)kg-1

300,000 d 62% 30% 500 Cmol(H+)kg-1

Fig. 11.1. Classification and chemical properties of groups of humic substances (after Stevenson and Elliott 1989)

Humic substances are generally classified into three main groups according to their solubility: – Humic acids are soluble in diluted bases and insoluble in acid medium; they result from precipitation by acidification of the alkaline extracts of soil. – Fulvic acids are soluble in both alkaline and acid mediums, the compounds remain in solution after precipitation of humic acids by acidification of the alkaline extracts of soil. – Humin is the humus fraction that cannot be extracted by acids or diluted bases.

Quantification of Humic Compounds

373

Other names have been given to certain humic substances as a function of their solubility in a range of different solvents. The best known characterize hymatomelanic acids as the fraction of humic acids which are soluble in ethanol. Among fulvic acids, an obsolete distinction is sometimes used to differentiate crenic and apocrenic acids. Some authors, for example Chamayou and Legros (1989), advise against the use of either term. Brown and grey humic acids can be distinguished. These two groups were first identified by Springer (1938) based on their flocculation or dissolution properties in a medium including different degrees of salts. Figure 11.1 shows some of the properties of these compounds. 11.1.2 Role in the Soil and Environment Humic compounds account for 60–70% of soil carbon, which itself represents the biggest reservoir of organic carbon on the earth’s surface. Humic compounds play a significant role as an atmospheric source of CO2 and as a carbon reservoir which is likely to react to the influence of different external factors (Schnitzer 1978). Stevenson (1982) quoted nine properties of humus with respect to their effect on the soil: – Its dark colour, which facilitates absorption of solar radiation and consequently warms the soil. – Its water-retention capacity; organic matter can to hold up to 20 times its own weight in water, thereby significantly improving the hydrous properties of some soils and particularly of sandy soils. – Its ability to combine with clay minerals resulting in the cementing of soil particles into structural units called aggregates thereby facilitating gaseous exchange and increasing permeability. – Chelation which forms stable complexes with many polyvalent cations and influences the availability of nutriments for plants. – Its solubility in water which is very reduced and the bonds with clays and certain polyvalent cations which minimize organic losses by leaching. – Its buffer effect, this effect appears at slightly acid, neutral and alkaline pH. – Its cation exchange capacity, 20–70 % of the cation exchange capacity of many soils is due to the presence of organic matter. – Its mineralization which releases CO2, NH4+, NO3–, PO43–, SO42– and represents a very significant source of nutrients for plant growth. – Its combination with other organic molecules which affects bioactivity, persistence and biodegradability of pesticides.

374

Organic Analysis

This chapter provides details of procedures for the main types of extraction and quantification of soil humic matter. The reference section at the end of the chapter lists a wider range of methods for use in this context. Extraction methods are described for whole soil samples prepared using standard techniques. However, it is often better to apply these methods to selected fractions after physical fractionation of the organic matter (cf. Chap. 9), in particular for the finest fractions. 11.1.3 Extraction Organic materials bond with polyvalent cations, hydroxides and clays to form organomineral complexes. The stability of these complexes varies considerably depending on the type of bond. Organic materials can be released by extractions that break down at least some of the organomineral bonds. Bruckert (1979) distinguished three types of extraction solutions: – Salt solutions can break electrostatic bonds by simple exchange of ions and help solubilize the organic molecules by ionizing the acid and phenolic functional groups; Bruckert proposed a sodium tetraborate solution at pH 9.7; the extracted organomineral substances of relatively weak molecular weight were referred to as mobile or easily available recently formed complexes (Duchaufour 1977); they are characterized by relatively low metal contents; tetraborate has no effect on calcium complexes. – Complex-forming solutions able to break the coordination bonds; the best known is sodium pyrophosphate solution which is generally used at pH 9.8; it breaks the bonds of complexes with metal sites on clays; it can also solubilize complexes with high metal contents (amorphous hydroxides), and dissolve calcium humates by forming a complex with calcium, but is as ineffective as tetraborate on allophanic complexes; all the extracted complexes are known as immovable. – Soda at pH 12 is in fact the most effective extractant as it is able to destroy most organomineral bonds and particularly those of the humic allophone–acid complexes of andosols. The standard techniques described in Sect. 11.2.1 concern: – Simple extraction using an alkaline solution, with or without a preceding acid attack. – Double extraction using a pyrophosphate solution followed by a soda solution (this method is used in the IRD laboratories, France).

Quantification of Humic Compounds

375

Section 11.2 evaluates the respective utility and precision of these methods and describes the main methods for the purification of extracted organic materials. Some alternatives to these extraction methods, including a technique for fractionation of the humin centrifugation pellet, are presented in Sect. 11.3. Techniques for the fractionation and characterization of humic compounds are described in Chap. 12.

11.2. Main Techniques

11.2.1 Extraction Principle As mentioned in Sect. 11.1.3, diluted soda solutions are the most powerful extracting reagents for humified matter. However, the use of these extraction solutions has been criticized for three main reasons (Bruckert 1979): – Neo-formation of soluble substances from non-humified plant materials. – Breakdown of humic substances by hydrolysis, oxidation or artificial polymerization. – Lysis of microbial organisms. Sodium hydroxide can destroy bacteria and empty them of their cytoplasmic contents, their cell walls then form a non-extractable residue. However, Schnitzer (1982) did not consider that extraction using diluted bases in a nitrogen atmosphere and at room temperature significantly modified the structure and characteristics of the extracted organic matter. Lévesque and Schnitzer (1966) showed that 0.1 mol L–1 soda solutions extract more organic matter than concentrated solutions. They also showed that 0.5 mol L–1 soda solutions extract organic matter with lower ash content. “Method Schnitzer (1982)” and “Method IHSS” were chosen to maximize extraction of humic compounds and to minimize their degradation. Extraction (one extraction only) is carried out (1) under nitrogen atmosphere, with a 0.1 mol (NaOH) L–1 soda solution described in “Method IHSS” (2) with the same reagent or 0.5 mol (NaOH) L–1 soda solution or pyrophosphate solution described in “Method Schnitzer (1982)”. “Method of Dabin (1976)” separates two types of extracted

376

Organic Analysis

compounds: organic materials extractable with a pyrophosphate solution at pH 9.8 and organic materials extractable later on with a 0.1 mol (NaOH) L–1 soda solution. Nègre et al. (1976) observed qualitative differences between the two extracts particularly in amino acid content; Thomann (1963) observed that pyrophosphate dissolves calcic humates by forming complexes with metal cations; an increase in pH acts more particularly on aggregate dispersion, and pH 9.8 corresponds to a stable stage in the curve of humus extraction as a function of pH. An alternative approach is acid pretreatment of the soil; this type of pretreatment facilitates the later extraction of humified matter by destroying carbonates and by solubilizing iron and aluminium hydroxides; however, the quantitative effect of the pretreatment is especially visible in calcareous soils. “Method IHSS” recommends systematic acid pretreatment with 1 mol (HCl) L–1 hydrochloric acid solution. “Method Schnitzer (1982)” recommends pretreatment with 0.05 mol (H+) L–1 hydrochloric or sulphuric acid solution only in the case of calcareous soils. “Method of Dabin (1976)” recommends systematic pretreatment with 2 mol L–1 phosphoric acid. This acid has two advantages (1) higher density (approximately 1.2) which is more favourable for the separation of light organic fragments (cf. Chap. 9), (2) it does not disturb wet carbon titration and thus enables quantification of the organic matter extracted by the acid itself (unbound fulvic acid). Equipment – Glass, polypropylene or polyvinyl extraction and centrifugation flasks (volume: 200 mL, and 300–500 mL) with screw caps, for use as centrifugation cylinders, capable of withstanding 10,000g. – Centrifuge (10,000g) equipped with rotor suitable for use with the centrifugation flasks. Products – Degassed inorganic water. Most commercial water is appropriate. It should first be checked for the absence of organic matter (blank assay corresponding to the type of characterization required). However, to eliminate organic matter from water, either (1) boil water for 2 h in the presence of 1% KMnO4 and H2SO4 then distil or (2) use deionized water purified on activated carbon (e.g. Millipore filter), then degas the water to eliminate dissolved oxygen in order to avoid oxidation of organic matter during extraction. Proceed either by boiling or by bubbling with nitrogen for 10 min.

Quantification of Humic Compounds

377

– 0.1 mol (NaOH) L–1 solution. Dissolve 8 g of soda pellets in a 2 L volumetric flask in degassed inorganic water, complete to 2 L, agitate and store in a carefully stopped bottle. – 0.5 mol (Na0H) L–1 solution. Same as above with 40 g soda for 2 L. – 10 mol (Na0H) L–1. Same as above with 400 g soda for 1 L. – 0.1 mol (Na4P2O7) L–1 solution. Dissolve 89.2 g of Na4P2O7,10H2O in degassed inorganic water, complete to 2 L and store in a carefully stopped bottle. – 0.1 mol (Na4P2O7,NaOH) L–1 solution. Dissolve 89.2 g of Na4P2O7,10H2O and 8 g of soda pellets in degassed inorganic water, complete to 2 L and store in a carefully stopped bottle. – 2 mol (HCl) L–1 solution. dilute 166.7 mL of concentrated HCl (d=1.19) in degassed inorganic water; complete to 1 L. – 6 mol (HCl) L–1 solution. Dilute 500 mL of concentrated hydrochloric acid in 1 L degassed inorganic water. – 1 mol (HCl) L–1 solution. Dilute 166.7 mL HCl in 2 L degassed inorganic water. – 0.5 mol (HCl) L–1 solution. Dilute 83.3 mL HCl in 2 L degassed inorganic water. – 0.05 mol (1/2H2SO4) L–1 solution. Dilute 27.8 mL of concentrated H2SO4 (d =1.81) in degassed inorganic water; cool and complete to 2 L. – 2 mol (H3PO4) L–1 solution. Dilute 136 mL of concentrated phosphoric acid (d=1.71) in degassed inorganic water and complete to 1 L. Procedures Method Schnitzer (1982) If the soil contains carbonates (reaction to diluted hydrochloric acid), leave it in contact with a 0.05 mol (H+) L–1 hydrochloric or sulphuric acid solution at room temperature until the end of gaseous emission. Rinse the excess acid with inorganic water and leave the soil to dry on a plate at room temperature. Weigh 10 g of air-dried soil in a 200 mL polypropylene flask. Add 100 mL of selected extraction solution (0.1 or 0.5 mol (NaOH) L–1, 0.1 mol (Na4P2O7) L–1 or mix 0.1 mol (Na4P2O7,NaOH) L–1. Purge the air out of the flask with a stream of nitrogen. Close carefully and agitate for 24 h at room temperature. Separate the dark supernatant solution from the solid phase by centrifugation (preferably for 10 min at 10,000g), suspend the residue in 50 mL degassed inorganic water, separate by centrifugation again and add the flushing water to the previous solution.

378

Organic Analysis

Acidify the alkaline extract to pH 2 with 2 mol (HCl) L–1 hydrochloric acid. Leave to stand for 24 h at room temperature then separate the soluble matter (fulvic acid) from the coagulated matter (humic acid) by centrifugation. The two fractions can be brought to dry by freeze-drying or by evaporation in a rotary evaporator at 40°C. Method IHSS1 Mix 20 g of air-dried soil with 1 mol (HCl) L–1 hydrochloric acid. Adjust to a pH of between 1 and 2 (15–20 mL soda 10 mol L–1) in such a way that the final volume of liquid is 200 mL (liquid/soil ratio = 10 mL per 1 g). Agitate for 1 h and separate the supernatant liquid by centrifugation. Neutralize the centrifugation pellet to pH 7 with 1 mol (NaOH) L–1 soda solution and add 0.1 mol (NaOH) L–1 solution under nitrogen atmosphere until a solution:soil ratio of 10:1 is obtained. Agitate for at least 4 h under nitrogen atmosphere. Leave to stand overnight and centrifuge. Acidify the centrifugation liquid to pH 1 with 6 mol (HCl) L–1 hydrochloric acid under agitation. Leave to stand for 12–16 h and centrifuge to separate the fulvic acids in solution from the coagulated humic acids. Method of Dabin (1976) Put 40 g of air-dried soil crushed and sieved on a 0.5 mm mesh sieve in a 300–500 mL centrifugation bottle. Add 200 mL of 2 mol (H3PO4) L–1 solution, agitate for 30 min with the back and forth shaker and centrifuge for 5 min at 1,500g. Filter the supernatant liquid on a flat filter in a 1 L glass bottle. Repeat the extraction two or three times in the same centrifugation bottle, filtering on the same filter and collecting the acid extracts in the same bottle. The filter contains light organic matter (LOM) from non-humified plant and animal residues (cf. Chap. 9); the acid solution contains a small fraction of organic materials called “free fulvic acids” (FFA) by Dabin (1976). Wash the centrifugation pellet two or three times with 200 mL inorganic water in the same bottle by agitating for 15 min; centrifuge and filter the washing water on the previously used filter to collect the light material still present in the washing water, discard the filtrate.

1

IHSS = International Humic Substance Society, Univ. of California, Los Angeles, CA 90024; Federal Center, mall stop 407, Box 25046, Denver, CO 80225.

Quantification of Humic Compounds

379

Add 200 mL of 0.1 mol (Na4P2O7) L–1 solution at pH 9.8. Agitate for 4 h on the back and forth shaker or leave in contact overnight agitating several times. Separate the supernatant liquid by centrifugation for 30 min at 3,000g and transfer it through a filter into a 1 L volumetric flask. Perform a second extraction in the same conditions and combine the extracts. If the second extract is dark in colour, perform a third extraction. Repeat a similar extraction sequence on the centrifugation pellet, with 0.1 mol (NaOH) L –1 soda solution instead of the pH 9.8 pyrophosphate solution. Humic acids of the pyrophosphate and soda extracts are separated from fulvic acid by acidification at pH 1 with the 2 mol (HCl) L–1 solution as described in “Method Schnitzer (1982)”. Ultimately, the following fractions are obtained: LOM, FFA, pyrophosphate fulvic acids (PFA), pyrophosphate humic acids (PHA), soda fulvic acids (SFA), soda humic acids (SHA), extraction residue or humin. Note In certain soils, the pyrophosphate and soda extracts can contain a high percentage of fine mineralogical clays (smaller than 0.2 µm). It is possible to separate these clays by flocculation with the addition of a little potassium sulphate, but there is a risk of simultaneously flocculating certain grey humic acids (cf. Sect. 11.2.4 and Chap. 12). After centrifugation, the flocculation pellet should either be titrated individually or combined with the previous humin pellet for carbon determination. 11.2.2 Quantification of the Extracts Principle Quantification is by carbon titration of each extract (cf. Sect. 11.2.1 above). The techniques used for carbon titration of whole soil can be used on the humin pellet (cf. Chap. 10). For LOM, it is preferable to use a combustion technique. The extracts can also be titrated by combustion on the residue after dry evaporation of an aliquot. Nitrogen can be measured in addition to carbon, hydrogen and possibly sulphur and oxygen by using an analyser of the CHN type. However, wet processes such as dichromate oxidation (described later) are often preferred for

380

Organic Analysis

titration of the extracts. Another technique calls for a titration apparatus with dissolved carbon. Many of these apparatus are based on titration of the carbon dioxide (generally by infra-red absorption) produced by oxidation of the solution with a powerful oxidant. The titration apparatus of dissolved carbon are rather expensive and reserved to precise environmental studies; the manufacturer’s instructions should be respected. In acid medium, dichromate oxidizes the carbon of the organic matter in CO2 according to the redox reaction: Cr2O72− + 6 e− + 14 H+ → 2 Cr3+ + 7 H2O C + 2 H2O → CO2 + 4e− +4 H+ With automated apparatuses, the quantity of CO2 released can be measured directly. With a traditional manual redox technique a dichromate excess is used, which is then back titrated by a ferrous iron solution: Fe2+ → Fe3+ + e− A mole of ferrous iron corresponds to 1/6 mol of K2Cr2O7 is 1/4 atom C or 3 g of carbon. Equipment – 50 and 100 mL Pyrex Beakers. – Precision burette (25 or 50 mL). – If necessary, a titration apparatus using carbon and dry combustion, but preferably a wet process. Reagents – 0.1 mol (Na4P2O7) L−1 and 0.1 mol (NaOH) L−1 solutions: see preparation in “Products” under Sect. 11.2.1. – Concentrated sulphuric acid (d=1.81). – 2 mol (½H2SO4) L−1 sulphuric acid: dissolve 56 mL of concentrated sulphuric acid (d=1.81) in 1 L inorganic water. – 0.1 mol (½H2SO4) L−1 sulphuric acid: 2.8 mL of concentrated sulphuric acid (d=1.81) in 1 L inorganic water. – 2% potassium dichromate solution: dissolve 20g of K2Cr2O7 in approximately 400 mL inorganic water, slowly add 500 mL of concentrated sulphuric acid, agitate, let cool and complete to 1 L with inorganic water. – 0.5 mol (1/ 6 K2Cr2O7) L−1 solution: gradually dissolve 24.52 g of potassium dichromate in inorganic water then complete to 1 L (solution for the titration of the Mohr salt).

Quantification of Humic Compounds

381

– 0.2 mol (FeSO4) L−1 Mohr salt solution: dissolve 78.4 g of Mohr salt (FeSO4,(NH4)2SO4,6H2O) in 500 mL water, add 20 mL of concentrated sulphuric acid, complete to 1 L. – Sodium fluoride in powder form. – Sulphuric diphenylamine solution: dissolve 0.5 g of diphenylamine powder in 100 mL of concentrated H2SO4, pour into 20 mL water, agitate and store in a brown bottle. Procedure Test samples Total organic materials of the soda and pyrophosphate extracts. In a 100–200 mL shallow beaker, put the exact volume of the extraction solution corresponding to 5–8 mg C. Calculate the volume of the aliquot on the basis of the total C analysis of the sample (cf. Chap. 10). The carbon of the total alkaline extracts is around 40% of total C; C extracted by pyrophosphate is about 25% of total C, and C extracted by soda (after pyrophosphate extraction) is about 15% of total C. Before titration, bring the sample to dry in a drying oven at 70°C. Humic acids. Take a sample of the extraction solution of more than 50% of the titration sample for total organic materials. Precipitate the humic acids at approximately pH 1 by adding 1 mol (H2SO4) L–1 (approximately 4–5 mL for 10 mL of total extract, 3 mL for 10 mL of pyrophosphate extract, 1.5 mL for 10 mL of soda extract); leave to flocculate for at least 4 h and centrifuge for at least 5 min at 3,500g; separate the supernatant liquid (fulvic acids) and wash with 0.1 mol (½H2SO4) L−1 solution. Dissolve the centrifugation pellet in 0.1 mol (NaOH) L–1 solution, place in a beaker and bring to dry at 70°C before analysis. Phosphoric acid extract. Take an exact aliquot of 100 mL; concentrate in the drying oven until approximately 10 mL remains (H3PO4 cannot go dry) and carry out titration on this concentrated solution. Fulvic acids. These can be titrated as total organic material after elimination of humic acids but are generally estimated by the difference between total organic materials and humic acids. Redox Titration After drying the sample specimens in the beakers, add 10 mL of the 2% potassium dichromate solution in sulphuric acid medium. At the same time, carry out a blank measurement with 10 mL of the same dichromate

382

Organic Analysis

solution in a beaker. Protect each beaker with a beaker cover and bring to a very gentle boil on a hotplate regulated at 215–220°C; Boil for 5 min but control boiling to avoid overheating or too much evaporation. Leave to cool, rinse the beaker cover and add in the beaker: 100 mL of inorganic water, 1.5 g of NaF (or 2.5 mL H3PO4) and three drops of diphenylamine solution. With a burette titrate with Mohr salt solution until the colour turns from purple to pale green. V and V’ volumes in mL of ferrous solutions are necessary for the respective titration of the blank and of the sample. If T is the concentration of the Mohr salt solution in mol (FeSO4) L–1, the quantity of Fe2+ equivalent to oxidant mobilized for the titration of the carbon of the sample is: T (V−V’) mmol(Fe2+), which according to the redox equations (cf. “Principle” under Sect. 11.2.2), corresponds to 3T (V −V’) mg of carbon in the beaker. With respect to the whole soil sample, the carbon content in mg (C) g−1 (soil) of the fraction is C = 3T(V−V’) Vt/(A mt) where: Vt: total volume of the humic extract in mL, A: aliquot of the humic extract used for titration in mL, mt: soil sample mass used for the extraction in g, T has to be determined by titration. Titration of the Mohr Salt Solution Put in a 250 mL beaker: – exactly 10 mL of the 0.5 mol (1/6 K2Cr2O7) L−1 solution – 100 mL of inorganic water – 15 mL of concentrated H2SO4 Let cool and add – 3.75 g NaF; – three drops of diphenylamine indicator Using the burette, titrate with the Mohr salt solution; if Vf is the volume in mL of the Mohr salt solution, the T of this solution will be T = 5/Vf mol (Fe2+) L–1. The carbon content in (11.1) can thus be expressed in mg (C) g–1 (soil) by: C = 15 (V–V’) Vt/(A mt Vf) (11.1’) (11.1)

Quantification of Humic Compounds

383

Remarks Ten millilitres of 2% dichromate solution corresponds to 20.4 mL of the 0.2 mol (Fe2+) L–1Mohr salt solution. The volume of Mohr salt solution used for titration of the sample should be between 7 and 15 mL; below this range, start the analysis again with a weaker sample; above this range start again with a stronger sample. The quantity of carbon in the phosphoric acid solution is usually low. In this case, oxidize with only 5 mL of 2% dichromate solution adding five drops of concentrated H2SO4 before boiling. Pyrex beakers are attacked by alkaline solutions and NaF in acid medium. They should be rinsed immediately after titration and reserved solely for this purpose. 11.2.3 Precision and Correspondence of the Extraction Methods Inter-Laboratory Study An inter-laboratory study by GEMOS 2 (Dabin et al. 1983) involved the quantitative comparison of the quantities of organic materials extracted on seven samples from soils from different areas of France: 1. A silt soil from the plates of Boigneville 2. A humocalcic soil from Pontarlier 3. A A1 horizon of a podzol from the forest of Villers Cotterets 4. A Bh horizon of the same podzol 3 5. A fersiallitic soil from near Montpellier 6. A gley soil with hydromull from Bonneveaux 7. A rendzina on chalk from Chalons sur Marne Each soil was analysed by four French laboratories: – the CIRAD3 laboratory of Montpellier – the pedology laboratory of the university of Poitiers – the IRD4 laboratory of Bondy – the pedology laboratory of the university of Besancon.

2

GEMOS = Groupe d’Etude des Matières Organiques des Sols, sub-group of the Association Française d’Etude des Sols (AFES), INRA, 78850 ThivervalGrignon, France. 3 CIRAD = Centre International de Recherche Agronomique pour le Développement, Avenue d’Agropolis, BP 5035, 34032 Montpellier Cedex, France. 4 IRD = Institute of Research for the Development (ex-Orstom), 32 Avenue Varagnat, 93143 Bondy, France.

384

Organic Analysis

The method of reference chosen for the comparisons was described in “Method IHSS”. In addition, the IRD laboratory in Bondy tested the method described in “Method of Dabin (1976)” on the same samples. Quantities Extracted by the IHSS Method The results of the inter-laboratory study are summarized in Fig. 11.2. Figure 11.2a shows the carbon value in g for 100 g of dry soil obtained by the sum of the three fractions: acid extract + alkaline extract + nonextracted residue, compared to total carbon measured on whole soil. The fact that the results are located close to the bisecting line shows the absence of bias between the two methods.

Fig. 11.2. Results of an inter-laboratory comparison using the extraction method described in “Method IHSS” (unpublished data). Results from four laboratories for the seven types of soils described in the text: (plus) soil 1, (open triangle) soil 2, (open square) soil 3, (open diamond) soil 4, (cross) soil 5, (filled diamond) soil 6, (open circle) soil 7

Quantification of Humic Compounds

385

Figure 11.2b shows the amount of carbon extracted with the hydrochloric acid solution compared to the total quantity extracted. The quantity of acid extracted was low, below or equal to 5% of total carbon. The two exceptions where the value was 10% are probably due to the presence of additional fragments of light non-humified organic matter. The values measured were very variable and there was no correlation with total carbon. Figure 11.2c shows the carbon of the alkaline extract compared to total carbon. Two remarks can be made: – in six of the seven soils, the soda solution extracted 20–40 % of total carbon. The values are more grouped than that of the acid extract. Their distribution revealed a maximum threshold of extraction for six samples located at approximately 40% of total carbon (line). The values located under this threshold are probably errors in measurement due to insufficient extraction. – there was an abnormally high value for soil 4; this can be explained by the fact the soil was from the deep Bh organic horizon of a podzol; the organic matter which was leached to this depth was much more soluble in the soda solution than in the six other soils; in this case, the reagent extracted more than 3/4 of the carbon of the sample. Figure 11.2d shows the quantity of carbon in humic acid forms compared to the carbon of the total alkaline extract. Whatever the type of soil, the carbon of humic acids accounted for approximately the 2/3 of the carbon of the alkaline extracts, and thus approximately 27% of total carbon. The deviation of the results around this value was limited. Precision of the IHSS Method Table 11.1 shows the results of an analysis of variance carried out on the data obtained with the IHSS method: are the measured values on each soil equal or significantly different compared to experimental error. The F test is the ratio of variance between soils (seven soils i.e. six degrees of freedom dof ) to within inter-laboratory variance (27 measurements or 20 dof ) represented by sr2. The pooled estimation of the standard error associated with a measurement from a given laboratory on an unspecified soil is indicated by “s ” in g (C) 100 g–1 (dry soil) in absolute values and by RSD (relative standard deviation) in relative values. These values representing the precision of an inter-laboratory reproducibility test are upper limits of error. Repeatability would be better within one laboratory with well-trained staff.

386

Organic Analysis

Carbon measurement was tested in the hydrochloric acid extract, in the soda extract, in the humin residue, in the sum of these three fractions and finally in humic acids and fulvic acids. Table 11.1 shows that: – It is impossible to control the quantities extracted by hydrochloric acid, the errors observed in this case being more significant than the variations between the soils; this confirms the distribution in Fig. 11.2b. – In all the other cases, the differences observed between the soils were significant compared to the residual error representing inter-laboratory variability.
Table 11.1. Precision of measurements in a comparative inter-laboratory test on seven soils using the IHSS method (F: test of significance of the soil values compared to the residual variance s 2r between the four laboratories (∗∗∗, significant difference between laboratory data at risk <1%, NS, no significant difference). s and RSD: expected absolute (g (C) 100 g–1dry soil) and relative (%) standard deviations in case of a measurement (no replicate) from any laboratory) Determination C- acid extract C- alcaline extract C- humin residue C- sum of fractions C- total soil C- humic acids C- fulvic acids F(6,21) 4.2 (NS) 107 ∗∗∗ 37 ∗∗∗ 83 ∗∗∗ 100 ∗∗∗ 67 ∗∗∗ 16 ∗∗∗ s2r 0.026 0.141 0.317 0.410 0.403 0.131 0.036 s 0.16 0.38 0.56 0.64 0.63 0.36 0.19 RSD% 83 17 22 13 13 22 35

– In the alkaline extracts, the precision of the measurement of humic acids is better than that in the fulvic acids; this is probably due to the greater abundance of humic acids. – The precision of the measurement of soil total carbon obtained by the sum of the first three measurements is of the same order of magnitude as that obtained by the direct measurement of carbon on the whole soil; moreover, the value obtained is in agreement with the rule of propagation of errors (Pansu et al. 2001); indeed, C-sum is obtained by: C-sum = C-acid extract + C-alkaline extract + C-humin residue

Quantification of Humic Compounds

387

In the case of normal laws: sC-sum = (s2C-acid extract + s2C-alcaline extract + s2C-humin)1/2 then: sC-sum = (0.026 + 0.141 + 0.317)1/2 sC-sum = 0.69 is very close to the values 0.64 and 0.63 found for the synthetic variable and the measurement of total carbon, respectively.
Comparison of the Methods Described in “Method IHSS” and “Method of Dabin (1976)”

12 C2 = C-sum of fractions of method of Dabin 10 8 6 4 2 0 0 2 4 6 8 10 12 C1 = C-sum of fractions of method IHSS C2=(1.01 + 0.05) C1 2 r = 0.99

Fig. 11.3. Comparison of the carbon rates found with the two methods of extraction tested. Averages of the results of four laboratories for the seven types of soils described in the text: (plus) soil 1, (open triangle) soil 2, (open square) soil 3, (open diamond) soil 4, (cross) soil 5, (filled diamond) soil 6, (open circle) soil 7

Figure 11.3 shows that the total quantities of carbon (acid extract + alkaline extract + humin residue) found with the two methods are very close. However, more detailed comparison of the data in Fig. 11.4 reveals analogies and differences between the methods: – Extraction of phosphoric acid described in “Method of Dabin (1976)” solubilizes between 1.2 and 4 times more (average 2 times more) organic matter than extraction with hydrochloric acid described in “Method IHSS” (Fig. 11.4a); the closest results for the two techniques were for the deep Bh horizon of the podzol. – Alkaline extraction gave quantitatively comparable results with the two methods (Fig. 11.4b) when the sum of the extracts pyrophosphate and

388

Organic Analysis

soda described in “Method of Dabin (1976)” are taken into account; the results are also comparable for humin carbon, although very slightly weaker in the method described in “Method of Dabin (1976)”, (Fig. 11.4d).

Fig. 11.4. Comparison of the fractions extracted with the two extraction methods in the comparative study. Averages of the results of four laboratories for the seven types of soils described in the text: (plus) soil 1, (open triangle) soil 2, (filled triangle) soil 3, (open diamond) soil 4, (cross) soil 5, (filled diamond) soil 6, (open circle) soil 7.

– On the other hand, there was a clear difference between the two methods with respect to the quality of the extracted organic matter since “Method of Dabin (1976)” provided significantly less humic acids; in Fig. 11.4c the C- “Method of Dabin (1976)”/C- “Method IHSS” ratio is approximately 2:3 and the behaviour of sample 4 (deep Bh horizon of

Quantification of Humic Compounds

389

the podzol) is a little different; the difference may be due to the effect of the extracting reagent on the extracted molecules i.e. polymerization in “Method IHSS”, macromolecular breakdown in “Method of Dabin (1976)” (cf. remarks in the introduction to this chapter). 11.2.4 Purification of humic Materials Introduction It is difficult to choose a precise procedure for purification, as many alternatives are possible. According to Schnitzer (1982), the prime objective of purification is to minimize the weight of ash, while the second objective consists in separating the organic molecules of lower molecular weight from the humic materials. However, these definitions may be insufficient because the modes of the bonds between humic, nonhumic and inorganic extracted materials are very complex and it is possible that many methods of purification have an influence on the structure of the final compounds isolated. Nègre et al. (1976) recommend dialysis (Viskins dialysis bags with 24 Å pores) to purify humic materials extracted by pyrophosphate. Like other authors before them, these authors noted some transformation of the fulvic acids into humic polymers of higher molecular weight. It is as if “during the dialysis, which is accompanied by a progressive return of the medium towards neutrality, the molecules that were depolymerized during the alkaline extraction could polymerize again by simple reestablishment of the CO–NH bonds, similar to the peptide bonds leading to the formation of the nucleic acids” (Nègre et al. 1976). Other simple purification methods enable elimination of certain minerals from the extracted solution by simple coagulation with the addition of a little sodium sulphate and centrifugation (Kumada et al. 1967) or ultracentrifugation (Jacquin et al. 1970). Humic acids can also be purified by dissolving them in soda medium then precipitating them again in acid medium (Lowe 1980) or by a second process of extraction–fractionation on extracts that have previously been freeze-dried (Schnitzer 1982), or by prolonged freezing of humic solutions which can fractionate the different phases (Bachelier 1983). The most effective procedure for purification – though only of humic acids – is to chemically attack the minerals with a diluted solution of the HCl–HF mixture to reduce the weight of ashes. Schnitzer (1982) noted that HCl-HF treatment can reduce the ash contents to less than 1%.

390

Organic Analysis

Jacquin et al. (1970) obtained less than 3% of ashes on three extracts purified with HCl–HF. Among the four purification methods tested by these authors, infra-red absorption spectra of purified humic materials showed that only the HCl–HF treatment almost completely eliminates the intense absorption bands of the phyllosilicates at 470, 520 and 1,030 cm–1 (Fig. 11.5). However, according to these authors, the hydrofluoric treatment leads to transformation of the chemical structure of the molecules, in particular a significant reduction in carboxylic acidity.

Fig. 11.5. Study by IR absorption spectrometry of the effect of three methods of purification on humic acids extracted from a deep Bh horizon of podzol (Jacquin et al. 1970): 1, non-purified humic acids; 2, humic acids purified by percolation on OH– and H+ resins; 3, humic acids purified by ultracentrifugation; 4, humic acids purified by HCl–HF treatment

Quantification of Humic Compounds

391

Fulvic acids can be purified in acid medium by adsorption on nonionic standard polyacrylic resins of the amberlite XAD-7 type (Aiken et al. 1979). After rinsing the resin in slightly acid medium to eliminate mineral salts, more than 98% of the fulvic acids can be recovered by elution at pH 6.5 (Gregor and Powell 1986). A simpler solution consists in eliminating the metal cations by repeated exchanges on a cation exchange resin in H+ form (Schnitzer 1982). Equipment – 150 mL Teflon or polypropylene bottles – Glass columns for chromatography with a diameter of 1 or 2 cm and a Teflon stopcock – Optional. Freeze dryer and freezer at −18°C Products – Dialysis bags with 24 Å pores – Non-ionic polyacrylic resin (Amberlite XAD-7 or similar) – Cation exchange resin Amberlite IR 120, or Dowex 50, in H+ form – HCl–HF mixture. Dilute 5 mL of concentrated HCl and 5 mL of 52% hydrofluoric acid in 990 mL inorganic water – Extracting reagents. see Sect. 11.2.1 Procedures Only the techniques for the purification of fulvic acids by cation exchange resins and techniques of purification of humic acids by HCl–HF attack are described here. However, as mentioned in “Introduction”, it is important to be very careful when choosing a purification method, which should take into account observations that need to be carried out on the purified products later on. For other methods of purification, see references in the last paragraph of “Introduction” under Sect. 11.2.4. Agitate a mixture of 1 g of humic acids and 100 mL of HCl–HF solution in a stopped polypropylene flask for 24 h at room temperature. Filter and suspend the filtrate again with 100 mL of HCl-HF solution. Repeat this treatment three or four times, carefully rinse the residue with inorganic water, then dry or freeze-dry. Purify the fulvic acid solution three or four times on cation ex-change resin in H+ form, then freeze-dry.

392

Organic Analysis

11.3. Further Alternatives and Complementary Methods

11.3.1. Alternative Methods of Extraction Though the methods described above using diluted soda and sodium pyrophosphate are the most widely used, alternative methods of extraction have been proposed. For example, Kumada et al. (1967) developed a method used at the University of Nagoya (Japan); Lowe (1980) used yet another technique at the University of British Columbia (Canada); and the two techniques were the subject of a comparative test by Lowe and Kumada (1984). Gregor and Powell (1986) developed a method for the extraction of fulvic acids with pyrophosphate in acid medium; this technique could avoids two potential problems: oxidation of the phenolic compounds in alkaline medium, and oxidation by the Fe3+ ions during acidification for precipitation of humic acids. Several comparative studies of extractions should also be mentioned. For example, Thomann (1963) compared 3% ammonium oxalate, 1% soda, 1% sodium fluoride and sodium pyrophosphate; Jacquin et al. (1970) compared soda, sodium pyrophosphate and ion exchange resins; Hayes et al. (1975) compared saline solutions, organic chelating agents, dipolar aprotic solvents, pyridine, ethylene diamine and soda in solution. They concluded that soda is the best of the reagents they tested for isolation of representative extracts of a broad range of humic substances. 11.3.2 Fractionation of the Humin Residue Principle Given that between 40 and 80% of total carbon cannot be extracted with alkaline solvents, Perraud (1971) and Perraud et al. (1971) proposed a technique for the fractionation of the non-extractable humin residue. They performed successively: – An alkaline extraction after two attacks with hot H2SO4: this extract was called “humin bound to iron” (Perraud et al. 1971); the sulphuric attack probably also releases sugars during the destruction of nonextractable polysaccharide like cellulose (cf. Chap. 13). – An alkaline extraction after six attacks with hot HF–HCl provided organic material bound to clays.

Quantification of Humic Compounds

393

However, these authors showed that the non-extractable fraction was still very high (37–52% of total carbon in their test) in spite of the total destruction of clays which was confirmed by X-ray. From 8 to 23% of this non-extractable fraction solubilized in CH3COBr probably corresponded to fresh or only slightly transformed organic matter that was not previously trapped in a mineral gangue. The fraction that remained insoluble in CH3COBr probably corresponded to either (1) organic matter very near to lignin but sufficiently transformed to be insoluble in acetyl bromide or (2) highly polymerized compounds in which the reduction of the functional groups probably resulted in their becoming insoluble in alkaline reagents. A procedure for fractionation of the humin residue developed from the method of Perraud et al. (1971) and used in IRD laboratories (Bondy, France) is described below. Procedure – Weigh 10 g of the extraction residue of Sect. 11.2.1 above that has been dried, crushed and sieved to 0.2 mm. – Add 50 mL of 1 mol (½H2SO4) L–1 sulphuric acid and heat at 70°C for 3 h. – Centrifuge at 3,000g for 15 min. – Wash twice with hot water and recover the centrifugation pellet. – Extract the centrifugation pellet with 50 mL 0.1 mol (NaOH) L–1 soda solution for 4 h with agitation and leave to stand overnight. – Centrifuge at 3,000g for 15 min. The supernatant liquid contains humin bound to hydroxides (HH); reserve for titration of fulvic and humic acids as described in Sect. 11.2.2; depending on soil type, the extract can be purified by flocculation of clays with a salt like sodium sulphate; add the flocculate to the centrifugation pellet. – Take an aliquot of the centrifugation pellet for intermediate titration of carbon and possibly of nitrogen; this titration enables identification of the fraction that is solubilized in the acid (e.g. polysaccharides) by calculating the difference. – On the other fraction of the centrifugation pellet (the larger fraction), destroy the clay particles by: (a) four successive attacks with 50 mL of the 1 mol (HCl–HF) L–1 mixture at 70°C for 3 h then centrifuge for 15 min at 3,000g (b) Attack with 50 mL of the 1 mol (HF) L–1 solution for 3 h at 70°C, centrifuge at 3,000g for 15 min, wash the centrifugation pellet with hot water and centrifuge again.

394

Organic Analysis

– Extract the centrifugation pellet with 50 mL of 0.1 mol (NaOH) L–1 solution agitate for 4 h and leave to stand overnight. – Centrifuge at 3,000g for 15 min. Recover the humic compounds of humin bound to the silicates (HS) from the solution; on these compounds measure the carbon of the fulvic and humic acids (cf. Sect. 11.2.2). – Inherited humin(IH), which is made up of small organic fragments resembling charcoal, can be recovered on the centrifugation pellet: (a) Add 50 mL H3PO4 d = 1.4; subject to ultrasound for 10 min, agitate mechanically for 30 min; (b) Centrifuge at 1,500g for 10 min, then filter the supernatant on a small funnel stopped with a glass wool plug. The carbon fragments recovered on this plug are the insoluble IH. As the liquids of previous acid attacks can also include suspended particles, after each centrifugation it is advised to filter the supernatants on the same glass wool plugs. After careful rinsing with inorganic water, dry the funnel at 50°C, crush the inherited humin and glass wool plug with an agate mortar and analyse carbon and nitrogen with a CHN analyser. The last centrifugation pellet is the final residue of nonextractable residual humin (RH), rinse twice with inorganic water to eliminate the phosphoric acid, dry, crush and analyse carbon and nitrogen with a CHN analyser. Calculations Data Collected Weight of soil sample at the beginning: Weight of humin (centrifugation pellet cf. “Procedures” under Sect. 11.2.1): Weight of the P1 sampling for sulphuric attack: Weight of the intermediate pellet (after HH extract): Weight of the P3 sampling for HF–HCl attack: Possible weight of other P3 sampling for C titration: (generally, P4 + P4’ = P3) Weight of non-extractable RH: Concentrations of the Extracts Humin (initial centrifugation pellet cf. “Procedures” under Sect. 11.2.1) Humin bound to hydroxides (mg C on initial humin) Intermediate centrifugation pellet (after HH extract) mg C g–1 Ch Chh Ci P0 P1 P2 P3 P4 P’4 P5

Quantification of Humic Compounds

395

Humin bound to silicates (mg C on the whole extracted) Carbon from the mixture of the glass wool and inherited humin (mg) Residual humin (mg C g–1) Calculations

Chs Cih Crh

Calculated on the initial soil sample, the carbon concentrations, (mg C g–1 dry soil) of humin before fractionation (H), humin bound to the hydroxides (HH), intermediate residue (I), humin bound to silicates (HS), inherited humin (IH), residual humin (RH) are expressed by: H = Ch P1/P0 HH = Chh P1 (P2 P0)−1 I = Ci P3 P1 (P4’ P0)–1 HS = Chs P3 P1 (P4 P2 P0)–1 IH = Cih P3 P1 (P4 P2 P0)–1 RH = Crh P5 P3 P1 (P4 P2 P0)–1 The H–(I+HH) value provides an estimate of the carbon dissolved in the hot sulphuric acid (polysaccharides of humin). The I−(HS+IH+RH) value provides an estimate of the carbon dissolved by the HF–HCl mixture.

References
Humic materials
Aiken GR McKnight DM Wershaw RL and MacCarthy P eds (1985) Humic substances in soil, sediment and water. I. Geochemistry, Isolation and Characterization. Wiley, New York, NY, 692 p Bruckert S (1979) Analyse des complexes organo-minéraux des sols. In : Pédologie 2.Constituants et propriétés du sol, Bonneau M and Souchier B ed., Masson, Paris, 187–209 Carter MR and Stewart BA eds. (1995) Structure and Organic Matter Storage in Agricultural Soils (Advances in Soil Science), Lewis Publishers, Inc., Boca Raton, FL

396

Organic Analysis

Chamayou H and Legros JP (1989) Les bases physiques, chimiques et minéralogiques de la science du sol. Presses Universitaire de France (Techniques vivantes), 608 p Duchaufour Ph (1977) Pédologie.1. Pédogénèse et classification. Masson, Paris, 477 p Flaig W, Beutelspacher H and Rietz E (1975) Chemical composition and physical properties of humic substances. In Soil Components, Vol.1 : Organic Components, Gieseking JE ed., 1–211 Hessen DO and Tranvik LJ eds. (1998) Aquatic Humic Substances : Ecology and Biogeochemistry (Ecological Studies, Vol .133). Springer, Berlin, Heidelberg, New York Kononova MM (1966) Soil Organic Matter – Its Nature, Its Role in Soil Formation and in Soil Fertility (translated from the Russian by Nowakowski PhD and Newman ACD), Pergamon Press, New York Magdoff F, Tabatabai MA and Hanlon E (1996) Soil Organic Matter: Analysis and Interpretation (Sssa Special Publication, No 46), American Society of Agronomy Piccolo A Ed. (1996) Humic Substances in Terrestrial Ecosystems. Elsevier, Amsterdam Schnitzer M and Kahn SU (1972) Humic Substances in the Environment. Dekker, New York, 327 p Schnitzer M (1978) Humic substances : chemistry and reactions. In : Soil Organic Matter, Schnitzer M and Kahn SU ed., Elsevier, Amsterdam, 1, 58 Springer U (1938) Der Heutige stand der Humusuntersuchungsmethodik mit besonderer Berücksichtigung der Trennung, Bestimmung und Charakterisierung der Huminsäùretypen und ihre Anwendung auf Charakteristische Humusformen. Zeitsc.f.Pflanzen, 6, 312–373 Stevenson FJ and Elliott ET (1989) Methodologies for assessing the quantity and quality of soil organic matter. In : Dynamics of Soil Organic Matter in Tropical Ecosystems, Coleman DC, Oades JM, Uehara G ed. University of Hawaii Press, Honolulu Hawaii 96822, 173–199 Stevenson FJ (1982) Humus Chemistry, Wiley, New York, 443 p Stevenson FJ (1994) Humus Chemistry : Genesis, Composition, Reactions, 2nd edition. Wiley, New York, 496 pages Tate RL (1992) Soil Organic Matter. Biological and Ecological Effects. Krieger, Melbourne FL, 304 p

Extraction, titration, purification and fractionation of humic materials
Aiken GR, Thurman EM and Malcolm RL (1979) Comparison of XAD macroporous resins for the concentration of fulvic acid from aqueous solution. Anal. Chem., 51, 1799–1803 Bachelier, G (1983) Variations périodiques dans le degré de condensation des acides humiques: mise en évidence par spectrofluorimétrie, corrélation avec la stabilité structurale des sols. Cahiers ORSTOM. Série Pédologie (FRA), 20(3): 247-254.

Quantification of Humic Compounds

397

Dabin B, Gavinelli E and Pelloux P (1983) Résultats de l’enquête analytique sur l’extraction et le dosage des matières humiques des sols-Réunion GEMOS Montpellier, Mai 1983. Document Orstom-Gemos, 16 p. multigr Dabin (1976) Méthode d’extraction et de fractionnement des matières humiques du sol, application à quelques études pédologiques et agronomiques dans les sols tropicaux. Cah.ORSTOM ser.Pédol., XIV, 287–297 Gregor JE and Powell HKJ (1986) Acid pyrophosphate extraction of soil fulvic acids. J. Soil Sci., 37, 577–585 Hayes MHB, Swift RS, Wardle RE and Brown JK (1975) Humic materials from an organic soil : A comparison of extractants and of properties of extracts. Geoderma, 13, 231–245 Jacquin F, Calvez C, Dormaar JF and Metche M (1970) Contribution à l’étude des processus d’extraction et de caractérisation des composés humiques. Bull.Ass.Fr.Etude du Sol, 4, 27–38 Kumada K, Sato O, Ohsumi Y and Ohta S (1967) Humus composition of mountain soils in central Japan with special reference to the distribution of P type humic acid. Soil. Sci. Plant Nutr., 13, 151–158 Lévesque and Schnitzer (1966) Effects of NaOH concentration on the extraction of organic matter and of major inorganic constituents from a soil. Can. J. Soil Sci., 46, 7–12 Lowe LE and Kumada K (1984) A comparison of two methods for routine characterization of humus in pedological studies. Soil Sci. Plant Nutr., 30, 321–331 Lowe LE (1980) Humus fraction ratios as a means of discriminating between horizon types. Can. J. Soil Sci., 60, 219–229 Nègre R, Ghiglione Cl, Pugnet T and Giraud M (1976) Influence des méthodes d’extraction et de purification sur la nature des acides humiques de la cédraie du Petit Lubéron. Cah.ORSTOM ser.Pédol., XIV, 337–350 Pansu M, Gautheyrou J and Loyer JY (2001) Soil Analysis – Sampling, Instrumentation and Quality control. Balkema, Lisse, Abington, Exton, Tokyo, 489 pp Perraud A (1971) La matière organique des sols forestiers de la Côte d’Ivoire., Thèse Docteur ès-sciences naturelles, Univ. Nancy I, 87 p. + annexes. Perraud A, Nguyen Kha, Jacquin F (1971) Essai de caractérisation des formes de l’humine dans plusieurs types de sols. C. R. Acad. Sci. Paris, série D, 272, 1594–1596 Schnitzer M (1982) Organic Matter Characterization. In : Methods of Soil Analysis, Part 2 Chemical and Microbiological Properties, 2nd edition, Page AL, Miller RH and Keeney DR ed. Agronomy monograph N°9, Am. Soc. of Agronomy, Madison, Wisconsin USA, 581–593 Thomann Ch (1963) Quelques observations sur l’extraction de l’humus dans les sols : méthode au pyrophosphate de sodium. Cah. ORSTOM ser. Pédol., 3, 43–72

12 Characterization of Humic Compounds

12.1 Introduction

12.1.1 Mechanisms of Formation The synthesis of humic substances has been the object of speculation for many years. Andreux (1994) distinguished lytic mechanisms (lysis of the cellular walls, proteolysis, ligninolysis, transformation of polyphenols and other organic components), the mechanisms of tanning and melanisation which include incorporation of nitrogen and oxygen (Maillard’s reaction: condensation of carbohydrates in the presence of amino nitrogen, condensation of polyphenol and amino acids in oxidizing medium) and the incorporation of inherited compounds. Schnitzer (1978) reported the following four hypotheses about the formation of humic substances: – Deterioration of plant material. Certain fractions of plant tissues, particularly woody materials, are only superficially decomposed in the soil to form humic substances; the nature of this “inherited humus” is thus strongly influenced by the nature of the original plant material; the first stage of humification provides the heaviest humic substances which can then be broken down into lighter substances and ultimately into CO2 and H2O. – Chemical polymerization. The plant materials break down into small molecules which are used as a source of energy and carbon by microorganisms; these micro-organisms then synthesize phenols and amino acids which are polymerized into humic substances; in this case the nature of original material has no effect on the type of substance formed.

400

Organic Analysis

– Cellular autolysis. The fragments resulting from autolysis of microbial and plant cells (amino sugars, acids, phenols and others aromatic compounds) condense and polymerize via free radicals. – Microbial synthesis. Microbes use plant tissue as a source of carbon and energy to synthesize intercellular organic materials of high molecular weight; at microbial death, these substances are released in the soil; they represent the first stage of humification and can then undergo extracellular microbial degradation into lighter molecules. 12.1.2 Molecular Structure Based on their solubility properties, humic substances are generally classified in the three following categories: humic acids, fulvic acids, humins (cf. Chap. 11).

Fig. 12.1. Structure of two humic macromolecules; only the nuclei are represented at a scale of molecular weight (mw) for (a) mw >50,000, and (b) mw <5,000. filled circle C=O, open circle CH, open square NH, ⎯ peptide bond, •⎯ OH carboxyl (Andreux, 1994)

In a review of the literature, Schnitzer and Kahn (1972) showed that these three humic fractions are structurally rather similar. The structure of the humic molecule can be described schematically as a nucleus rich in hydroxy-quinonic units linked by C–C or C–O–C bonds (Andreux 1994). The frequency of distribution of the proteic and polypeptide chains fixed

Characterization of Hmic Compounds u

401

on this nucleus varies with the type of soil and the precursors. The molecular size of these structures (Fig. 12.1) depends both on the dimension of the nucleus and on the nature of the phenolic precursors. Their affinity for aqueous solvents thus depends on the number and the length of the peripheral hydrophilic chains carrying the free –COOH groups of polypeptides, however, with significant acidity of the matrix (Andreux 1994). The properties of the three humic fractions differ especially with respect to their molecular weight, their ultimate analysis, and the number of functional groups. Fulvic acids have the lowest molecular weight, contain more oxygen but less carbon and nitrogen than the other two. The functional groups containing oxygen (CO2H, OH, C=O) have a rate per unit of weight that is higher in fulvic acids than in humic acids and humin (Stevenson 1982). Humic acids are also extractable from charcoal (Lawson and Stewart 1989) and research using 13C nuclear magnetic resonance spectroscopy (cf. Sect. 12.3.4) has shown that the stable carbon of Australian soils is mainly charcoal (Skjemstad et al. 1996). Although structural models of humic molecules have been proposed (e.g. Schulten 1995, Schulten and Schnitzer 1997, Schulten and Leinweber 2000, Schulten 2002, Lodygin and Beznosikov 2003), most of the concepts concerning molecular structure are not applicable to humic acids (McCarthy 2001). Further research is needed to obtain precise information about these molecules, their bonds with inorganic (Schulten and Leinweber 2000) and xenobiotic (Piccolo et al. 1999) molecules and their distribution in different soil types, which, in turn, will improve understanding of their role in the environment. All modern instrumental methods are needed for these researches (Hatcher et al. 2001, Piccolo and Conte 2003).

12.2 Classical Techniques

12.2.1 Fractionation of Humic Compounds Principle For many years humic acids were characterized according to the degree they bonded with clays (Springer 1938, Tiurin 1951, Duchaufour 1954, 1956).

402

Organic Analysis

Brown or free humic acids were considered to be only slightly bonded to clays, to have the lowest molecular weight and to be relatively insensitive to the flocculating action of electrolytes; they were said to derive from the oxidation of lignin under the action of the polyphenoloxidase and to characterize acid soils in particular, e.g. forest soils in a wet climate (Duchaufour 1956). Grey humic acids are darker in colour, form closer bonds with mineral colloids, have more condensed molecules, flocculate easily in the presence of electrolytes; this type of humus is characteristic of black soils, but is also relatively abundant in all soils that are rich in calcium. Early methods of fractionation of humic acids were thus based on the properties of the two main types of compounds: extraction in the presence of a flocculating agent in the case of brown humic acids followed by washing the residue in water in the case of grey humic acids (Duchaufour 1956), direct soda extraction of brown humic acids then extraction of two fractions of grey humic acids as a function of their bonds with calcium or iron and aluminium (Duchaufour 1957). To simplify the procedure, Duchaufour and Jacquin (1966) proposed a method with electrophoretic fractionation of humic acids on pyrophosphate extracts (cf. “Method of Dabin (1976)” Chap. 11). This method was compared with that of Tiurin (1951) by Dabin and Thomann (1970); it was widely used in France, in particular in IRD1 laboratories (Ratsimbazafy 1973, Dabin 1980). This is the first fractionation method described later. The technique of fractionation of humic acids by exclusion chromatography (Bailly and Margulis 1968, Bailly and Tittonel 1972) is also described later, along with a fractionation procedure for fulvic acids. Other complex techniques for fractionation of humic compounds are described more briefly in Sect. 12.3.1. Equipment – Plastic electrophoresis tank with three compartments (Fig. 12.2) – stabilized supply of direct current adjustable between 0 and 600 V – photoelectric densitometer to read the electrophoresis diagrams – columns for liquid chromatography, diameter: 2.5 cm, length: 80 cm – UV–visible detector equipped with a recorder or a system for data acquisition – fraction collector (optional) – peristaltic pump with flow of 50 mL h–1 (optional).

1

IRD = Institute of Research for Development (ex-Orstom), Bondy, France.

Characterization of Humic Compounds

403

Products – Filter paper tapes (Arch 302 or Whatman no. 1 or 2), 5 cm in width and 35 cm in length, cut perpendicular to the direction the filter is filled (when a square sheet of paper is held by one side, the direction perpendicular to filling is where the paper curves most under its own weight) – extraction solutions (see Sect. 11.2.1 in Chap. 11) – buffer solution for electrophoresis: in a 2 L volumetric flask add 13.6 g of monobasic potassium phosphate, 2.5 g of soda pellets and approximately 1.5 L of inorganic water (cf. Sect. 11.2.1 of Chap. 11), adjust pH to 7.4 with soda if necessary, agitate well and complete to 2L – standard dextrane gel Sephadex G25 for molecular weights ≤5,000 – standard dextrane gel Sephadex G75 for molecular weights ≤50,000 – standard dextrane gel Sephadex G200 for molecular weights ≤200,000 – polyvinylpyrrolidone

Fig. 12.2. ouan tank for paper electrophoresis of humic acids J

– TRIS buffer (pH 9, ionic force 0.5): mix 414 mL of M 2-amino-2 hydroxymethyl-propane-1, 3-diol, 50 mL of 1 mol (HCl) L–1solution and complete to 1 L inorganic water – Borax buffer (pH = 9.1, ionic strength = 0.075): 0.025 mol (Na2B4O7) L–1.

404

Organic Analysis

Procedure for Electrophoresis Fractionation of Humic Acids – Take a quantity of humic extract solution (pyrophosphate or soda) corresponding to 25 to 50 mg carbon (cf. titration in Sect. 11.2.2 of Chap. 11). – Precipitate the humic acids at pH 1 with sulphuric acid, centrifuge and wash the centrifugation pellet several times with a 1 mol (½H2SO4) L–1 solution. – Dissolve the humic acids in approximately 1 mL of normal soda solution so as to obtain a rather thick solution but without solids (cf. purification in Sect. 11.2.4 of Chap. 11) and with a homogenous concentration of carbon; store in a well-stopped hemolysis tube. – Fill the two external compartments of the electrolysis tank to the same level with electrophoresis buffer solution; wet the paper strip in the buffer solution, dry it between two sheets of filter paper and place it in the electrophoresis tank, pull it tight between the two external compartments as shown in Fig. 12.2. – With a 100 mm3 micropipette deposit approximately 40 mm3 of humic solution at a distance of 5 cm from the cathode following a straight line down the centre of the strip of paper and leaving 1 cm free on each side of the paper; in the case of very concentrated solutions, the deposit can be reduced to 20 mm3. – Place the lid on the tank and start the electrical current; regulate the current at 10 V per cm of paper, or approximately 200 V; the intensity of the current depends on the number of paper strips for simultaneous electrophoresis and on the conductivity of the electrolyte (approximately 15 mA with four strips). The negatively charged humic molecules migrate towards the anode; the smaller the molecules, the faster they migrate (e.g. brown humic acids, BHA). Migration is rapid at the beginning then slows down. The time needed for standard electrophoresis is 3 h which corresponds to 10–12 cm displacement by brown humic acids. – Shut off the current and rapidly remove the paper strips, place them on a flat surface and dry them under IR radiation or in the drying oven at 60°C. Record the radiation transmission of each strip of paper with a photoelectric densitometer. – An electrophoresis diagram (Fig. 12.3) is obtained in which: (1) grey humic acids (GHA) often display a distinct narrow peak up to 1 cm from the starting line; in the case of the chernozems, Duchaufour and Jacquin (1963) reported a second peak which could migrate up to 2 cm from the starting line; in tropical soils, GHA are frequently spread

Characterization of Humic Compounds

405

out over 3 or even 4 cm with either a single peak, or two or more peaks; it is sometimes rather difficult to detect the limit of GHA and their peak limit has consequently been arbitrarily fixed at 1/3 of the overall length of the diagram, i.e. 3–4 cm for a length of 9–12 cm

Fig. 12.3. Examples of electrophoresis diagrams of humic acids extracted with a pH 9.8 pyrophosphate solution in a range of tropical soils (Dabin 1980). On the left: direct readings with the optical densitometer at wavelengths of 512 nm (top curve) and 625 nm. On the right: corresponding integral curves giving in the y-coordinate surfaces corresponding to grey humic acids (GHA: from 0 to 1/3 of the graph), intermediate humic acids (IHA: between 1/3 and 1/2), brown humic acids (BHA: more than ½ of the graph). a: tropical podzol (histic tropaquod), A horizon, b: tropical podzol (histic tropaquod), B horizon, c: humiferous ferrallitic soil (umbriorthox), surface horizon, d: humiferous ferrallitic soil (umbriorthox), subsurface horizon, e: weathered tropical soil (oxictropudalf).

(2) the humic compounds located between 1/3 and 1/2 of the length of the diagram are called intermediary humic acids (IHA)

406

Organic Analysis

(3) the brown humic acids (BHA) are spread out between the middle and the other end of the diagram. – The surface area of each fraction is determined; SG is the surface of the GHA, SI that of the IHA, SB that of the BHA, ST is the total surface area and % qt is the total quantity of humic acids; the total quantity of each fraction is expressed by: GHA % = qt SG/ST IHA % = qt SI/ST BHA % = qt SB/ST Figure 12.3 shows some examples of electrophoresis diagrams obtained with this method. Remarks Any liquid circulating on the paper strips can also cause migration of humic acids. This should be avoided by making sure that the liquid is at the same level at the anode and cathode (otherwise there is a risk of siphoning by the paper), and by avoiding evaporation on the paper. Evaporation can cause the liquid to move by aspiration on both sides of the paper strip. Thus, even without electrical current, certain humic acids may migrate and be found in the centre of the strip. With electrical current, migration will stop when the speed of migration due to the electric potential is equal to the speed of the liquid circulating in the opposite direction. It is thus imperative to use a lid to limit evaporation; the lid should be shaped like a roof to prevent drops of condensation falling onto the paper strips. The passage of the current enriches the cathodic compartment in the soda and raises its pH; in the case of several electrophoresis series this phenomenon can be compensated for by reversing the direction of the current and thus of the deposit. Procedure for Humic Acid Fractionation on Dextrane Gels Fractionations can be carried out in a simple way with elution using inorganic distilled water (Bailly and Margulis 1968, Bailly and Tittonel 1972): – prepare a concentrated solution of humic acids in the same way as for electrophoresis but with a sample specimen corresponding to 2–10 mg carbon, diluted in 3–10 mL of 1 mol (NaOH) L–1 solution – fill the columns with the Sephadex gel: 50 cm for G25 and G75, 70 cm approximately for G200, these heights are likely to vary with the type of humus (Bailly and Tittonel 1972)

Characterization of Humic Compounds

407

– deposit the test specimen on the G25 column and elute with inorganic water either by gravity (descending chromatography), or with a peristaltic pump at a flow of approximately 30 mL h–1 (ascending chromatography) – carry the effluent in a UV photometer detector regulated at 253.7 nm; record the chromatogram and collect the fractions corresponding to the main peaks (Fig. 12.4);

Fig. 12.4. Examples of separation by exclusion chromatography on dextrane gel (Bailly and Tittonel 1972). On the left, grassland podzolic soil, A1 horizon, 0–6 cm depth; on the right, forest grey soil, A2 horizon, depth 16–29 cm. a: fractionation of the humic acid extract on Sephadex G25 F, test specimen corresponding to 4.7 mg C, b: fractionation on Sephadex G75 of peak V obtained in a, c: fractionation on Sephadex G200 of peak VIII obtained in b.

– the first eluted peak corresponds to the largest molecules that were not separated by gel exclusion; collect this fraction and subject it to exclusion chromatography in a similar way but on column G75; fractionate the new first peak eluted on this column again on column

408

Organic Analysis

G200; Fig. 12.8 presents two series of examples of chromatograms obtained successively on the three types of columns with two soils. This type of fractionation by gel permeation can also be performed in buffered solutions to avoid interactions between the gel and the solution, e.g. Tris and Borax buffers prepared in the same way as in “Products” in Sect. 12.2.1 (Cameron et al. 1972a). Fractionation of Fulvic Acids The method of Lowe (1975) is a very simple way to separate fulvic acids into two fractions: a coloured polyphenolic fraction and an almost colourless fraction with a prevalence of polysaccharide. Since Ch, Cf, Ca are carbons in humic acids, fulvic acids and their polyphenolic coloured fraction respectively, the Ch:Cf and Ca:Cf ratios were linked to the types of horizons used in the Canadian soil classification system which facilitate the distinction between some of these types (Lowe 1980): – treat a fraction of the fulvic acid extract using 1 g polyvinylpyrrolidone for 100 mL of solution – agitate intermittently for 30 min and filter – titrate carbon on the filtrate (cf. Sect. 11.2.2 of Chap. 11) to quantify the Ca fraction; the fulvic acid sample must be sufficient in volume to allow titration with acceptable precision. 12.2.2 Titration of the Main Functional Groups Principle The procedures described here are based on the measurement of total acidity and of carboxylic acidity of Wright and Schnitzer (1959) and of Schnitzer and Gupta (1965). For measurement of total acidity, the sample is treated with a barium hydroxide solution under N2 for 24 h. The Ba(OH)2 remaining in the solution after the reaction is then back titrated with a standard acid solution. For the titration of carboxylic groups, the humic materials are agitated for 24 h with calcium acetate solution in excess which causes the release of acetic acid according to a reaction of the type: 2 RCOOH + (CH3COO)2Ca → (RCOO)2Ca + 2 CH3COOH

Characterization of Humic Compounds

409

The acetic acid released is then titrated with a standard soda solution. The proportion of phenolic groups is calculated by the difference between total acidity and the acidity of the carboxylic groups. Other measurement techniques for functional groups (phenolic, alcoholic, ketonic, quinoid) are described briefly in Sect. 12.3. Equipment – 125 mL Erlenmeyer flasks with screw caps – titrimeter equipped with a combined electrode for the measurement of pH. Products – 0.2 mol (1/2 Ba(OH)2) L–1 barium hydroxide solution: weigh 31.548 g of Ba(OH)2,8H2O (quality containing the minimum of carbonate), dissolve in inorganic CO2–free water, complete to 1 L and protect from atmospheric CO2 with a trap containing soda lime – 0.5 mol (HCl) L–1 hydrochloric acid solution: prepare using standard commercial dose; dilute with inorganic CO2–free water. – 1 mol (1/2Ca(CH3COO) 2) L–1 calcium acetate solution: dry the pure product at 105°C and weigh 79.085 g under anhydrous atmosphere, dissolve in inorganic water and complete to 1 L. – 0.1 mol (NaOH) L–1 soda solution: prepare using standard commercial dose; dilute with inorganic water, complete to 1 L and protect with a trap containing soda lime during storage. Procedure Total Acidity To obtain the maximum number of exchangeable sites in acid-active form, it is advisable to work with carefully purified humic materials in order to reduce their ash rate (cf. Sect. 11.2.4 of Chap. 11). Place an exact weight of between 50 and 100 mg of freeze-dried humic material in a 125 mL Erlenmeyer flask with a screw cap and add exactly 20 mL of 0.2 mol (Ba(OH)2) L–1 solution. Perform a blank assay containing only the 0.2 mol (Ba(OH)2 L–1 solution without the sample. Put the flasks under nitrogen atmosphere, stop well and agitate for 24 h at room temperature. Filter the suspension, wash the residue well with distilled CO2–free water, add the washing waters to the filtrate and carry out potentiometric titration of the resulting extraction solution with 0.5

410

Organic Analysis

mol (HCl) L–1 solution up to pH 8.4. Vb and V are the volumes of standard acid solution for titration of the blank assay and of the sample, respectively, Na is the acid normality and P the weight of the sample (mg); total acidity At in milliequivalents per gram of humic material is expressed by: At=1,000(Vb Vs) Na/P Carboxyl Groups Place in a 125 mL Erlenmeyer flask with a screw cap an exact weight of between 50 and 100 mg of humic material; add 10 mL of 1 mol ½ Ca(CH3COO)2 L–1 solution and 40 mL of inorganic CO2–free water. At the same time, perform a blank assay containing the reagents without the humic sample. After 24 h of shaking at room temperature, filter the suspension, rinse the residue with distilled CO2–free water, combine the filtrate and washing water and perform potentiometric titration with the 0.1 mol (NaOH) L–1 solution up to pH 9.8. If Vs and Vb are the volume of titrating solution for the sample and blank assay, respectively (mL), Nb is the normality of the standard soda solution (mol L–1), P the sample weight (mg); the carboxylic acidity Ac in mol (COOH) g–1 humic material, is expressed by Ac = 1,000(Vs Vb) Nb /P. Phenolic Acidity This Ap acidity can be expressed in mol (phenolicOH) g–1 humic material by Ap = At –Ac. 12.2.3 UV–Visible Spectrometry Principle Among other molecular spectrometry techniques, UV spectrometry records electronic energy transitions in the molecules whereas lowerenergy infra-red radiation records variations in molecular kinetic energy. Although absorption in the ultraviolet and visible field of the electromagnetic spectrum does not give a band that is very characteristic of humic compounds (Schnitzer and Kahn 1972, Schnitzer 1978), the E4:E6 ratio of absorbance at 465 (E4) and 665 nm (E6) is often used to characterize humus. Ratios lower than 5 are characteristic of humic acids

Characterization of Humic Compounds

411

while fulvic acids have higher ratios; this ratio is independent of the concentration of humic materials but is not the same for humic materials extracted from different types of soils.

Fig.12.5. Effect of certain factors on the value of the E4:E6 ratio (Chen et al., 1977): (a) effect of the pH in different fractions of a fulvic acid (Bh horizon of a Canadian podzol) separated on Sephadex gel; the molecular weights measured by osmometry (Hansen and Schnitzer, 1969) were: 883 for AF-IV, 1181 for AF-III, 1815 for AF-II, 2110 for AF-I and >2110 for >AF-I, (b) effect of pH in some humic acids of Ah horizons of Canadian soils from the area round Alberta: Black Chernozem, Solod, Solonetz, (c) effect of the concentration of free radicals measured by electron spin resonance (cf. Sect. 12.3.6) for the fulvic acid used in a, (the numbers on the curve are pH values).

Kononova (1966) believed that the E4:E6 ratio was related to the degree of condensation of the aromatic carbon lattice, a weak ratio indicating a high degree of condensation of the aromatic humic

412

Organic Analysis

components, a strong ratio indicating the presence of a higher proportion of aliphatic structures. Chen et al. (1977) made a thorough study of the information provided by the E4:E6 ratio. Their study showed that the E4:E6 ratio: – is mainly governed by molecular size (or molecular or particle weight; Fig. 12.5a, b) – is highly affected by the pH (Fig. 12.5a, b) – is correlated with the concentration of free radicals (Fig. 12.5c), and with O, C, CO2H contents and total acidity (these measurements are also correlated with the size of the particles) – does not present a direct correlation with the concentration of condensed aromatic nuclei, which would invalidate the assumption of Kononova (1966) – is independent of the concentration of humic or fulvic acid, at least in the field of 100–500 ppm, which confirms the other assumption of Kononova. In agreement with Kononova (1966), these authors finally showed that the most favorable range of pH to measure E4:E6 ratios is between 7 and 8. This can be obtained by dissolving humic material in a 0.1 mol (NaHCO3) L–1 solution at a concentration of between 200 and 400 mg kg–1. Ghosh and Schnitzer (1979) proposed a mechanism linking the macromolecular characteristics of humic substances and UV–visible absorption: optical density decrease with an increase in the concentration of neutral electrolyte, indicating a reduction in the size of the particles probably due to a rolling up of the macromolecule. Equipment – UV–visible spectrograph with an adjustable double beam and fixed wavelength (465 and 665 nm) or preferably with variable wavelength between 200 and 700 nm – quartz tanks for UV spectrometry. Reagents – 0.05 mol (NaHCO3) L–1 solution: in a 1 L volumetric flask dissolve 4.200 g of NaHCO3 (quality suitable for spectrography) in inorganic distilled water, complete to 1 L and stop the flask well before storage.

Characterization of Humic Compounds

413

Procedure Dissolve 2–4mg of humic material in 10 mL of the 0.05 mol (NaHCO3) L–1 solution. Check the pH after dissolution, it should be close to 8 (the pH of a 0.05 mol (NaHCO3) L–1 aqueous solution is 8.3). Fill the quartz measurement tank to mid-height with this solution and fill the reference tank with a pure 0.05 mol (NaHCO3) L–1 solution. Measure the absorbance at 465 and 665 nm. The ratio of these two absorbencies is the E4:E6 ratio. If more detailed studies are required, the optical density (OD) spectrum can be recorded in the 200–350 nm UV range (Ghosh and Schnitzer 1979) or in the 400–700 nm range; in the latter case, the straight lines Log(OD) = f (Log k) can be plotted and, according to Chen et al. (1977), its slope should be equal to –6.435 Log(E4:E6) 12.2.4 Infra-Red Spectrography Principle The infra-red spectrum between 1 and 100 µm wavelength makes it possible to observe the vibrations of stretching and deformation of the molecules (as the spectrum of molecular rotation corresponds to less energetic radiations of wavelengths higher than 100 µm). In practice, the most useful spectral field for organic chemistry is in medium IR between the two wavelengths λ of 2.5 and 15 µm corresponding to wavenumbers 1/λ between 4,000 and 660 cm–1 (cf. Sect. 5.1.1 of Chap. 5). The near IR zone can also be widely explored with the help of chemometrical software (cf. Sect. 5.3.1 of Chap. 5). In humic substances, the IR spectrum mainly reflects oxygenated functional groups such as –CO2H, –OH and C–O. Some IR bands are particularly well defined (Schnitzer 1971) at wavenumbers 3,400 cm–1 (H bound to OH), 2,900 cm–1 (aliphatic CH bonds), 1,725 cm–1 (C–O of CO2H, C–O elongation of ketonic carbonyls), 1,630 cm–1 (aromatic C–C, H linked to C–O of carbonyls, COO–), 1,450 cm–1 (aliphatic CH), 1,400 cm–1 (COO–, aliphatic CH), 1,200 cm–1 (CO stretching, OH deformations of CO2H), 1,050 cm–1 (Si–O of the silicated impurities). The IR spectrum does not provide much information on the chemical structure of the core of humic substances. However, it is very useful for preliminary characterization of humic materials of different origin (Fig. 12.6) to determine the effect of different extractions or chemical

414

Organic Analysis

purification agents (cf. Sect. 11.2.4 of Chap. 11), and to study the reaction of derivatisation such as silylation, methylation, and acetylation. The IR spectrum also makes it possible to detect changes in the structure of humic materials following oxidation, pyrolysis or other treatments. Lastly, it is a practical method to characterize the formation of metal–humate and clay–humate complexes or to detect interactions between humic materials and other organic molecules such as pesticides.

Fig. 12.6. nfra-red spectra of some fulvic (on the left) and humic (on the right) I acids of standard and reference samples of IHSS (Senesi et al. 1989): (a) Suwannee River Standard, code IHSS 1S101F (AF) and 1S101H (AH), (a1) Suwannee River Reference, code IHSS 1R101F (AF ) and 1R 101H (AH), (b) Nordic aquatic, code 1R105F and 1R105H, (c) Soil St, code 1S102Fand 1R102H, (d) Peat, code 1R103F and 1R103H, (e) Leonardite, code 1R 104 H.

Great care should be taken in interpreting the spectra and particularly to avoid confusing the organic or mineral origin of the absorption bands (Russel and Anderson 1977). Equipment – Double beam IR absorption spectrograph with a field frequency between 300 and 4,000 cm–1

Characterization of Humic Compounds

415

– manual hydraulic press for the preparation of pellets for IR spectrography (e.g. standard 12 ton Spex-Carver) – polished stainless steel vacuum pelletizer, diameter 13 mm. NB: it is also possible to work without a pelletizer or a hydraulic press; spectra can be obtained in suspensions that are maintained between two IR-transparent blades; spectra can also be obtained in solutions with a solvent that is transparent to IR. The technique described below is not the cheapest but has been shown to be particularly suitable for IR spectrometry. Product – Potassium bromide in powder form for IR spectrometry. Procedure With the agate mortar, prepare a KBr pellet by mixing 1 mg of humic material with 400 mg of dry KBr. Place the powder in the pelletizer, put it under vacuum and press with a pressure of 7,600 kg cm–2 for 20 min (cf. “Preparation of Discs (Solid Solution)” in Chap. 5). Unmould the pellet, which should be vitrified, and place it in the measuring cell of the IR spectrometer, put a pellet of pure KBr in the reference cell and record the spectrum between 300 and 4,000 cm–1.

12.3 Complementary Techniques

12.3.1 Improvements in Fractionation Technologies The techniques described later are the result of improvements in electrophoresis and gel exclusion chromatography (cf. Sect. 12.2.1). Electrophoresis – Electrofocusing Method Cacco et al. (1974) proposed an improvement of the electrophoresis of humic compounds by using the electrofocusing method described by Righetti and Drysdale (1971). In this technique, humic compounds migrate from anode to cathode in a polyacrilamide gel in the presence of ampholines which cause a pH gradient. The migration stops when each compound reaches its isoelectric point. Figure 12.7 shows the results of

416

Organic Analysis

isoelectrophoretic characterisation (Cacco and Maggioni 1976) of fulvic and humic acids extracted with pyrophosphate at pH 7 from an alpine podzol. Rusina et al. (1983) suggested a system of calculation of molecular parameters by electrophoretic mobility in the polyacrylamide gel. Electrophoresis techniques are generally used for the study of molecular size but also of the electrical charge of humic substances (Duxbury 1989).

Fig. 12.7. Isoelectrophoretic characterization of fulvic (deep Bh horizon) and humic (surface A1 and deep Bh horizons) acids extracted from an alpine podzol (Cacco and Maggioni 1976)

Gel Exclusion Chromatography The gel can be calibrated with the molecular weights of humic acids measured jointly using other techniques. Figure 12.8 shows the calibration curves obtained by Cameron et al. (1972a) with four types of gel; the molecular weights (on the x-coordinate) were measured by a sedimentation technique using ultracentrifugation (Cameron et al. 1972b). In such studies, the gel is characterized by the Kav parameter suggested by Laurent and Killander (1964): Kav = (VR Vo)/(Vt Vo), Where is the VR: volume of retention, VO: volume of pores and Vt: volume of total column.

Characterization of Humic Compounds

417

The median values of Kav are adjusted to the median molecular weights M (Fig. 12.8) by means of two constants k1 and k2 according to law: Kav = k1 ln M + k2.

Fig. 12.8. Calibration of four types of gel (Cameron et al. 1972a) for the fractionation of humic acids in comparison with other macromolecular compounds (solid lines and solid circles: humic acids, dotted lines and open circles: data from the protein fractionation, sucrose is used to determine the stationary liquid volume Vt See text for Kav and M: (a) Sephadex G100 in tris buffer (cf. Sect. 12.2.1 for preparation), (b) Sephadex G100 in borax buffer (cf. Sect. 12.2.1 for preparation), (c) Biogel P-150 in tris buffer and (d) Sepharose 6B in tris buffer.

Nowadays, fractionation by gel permeation can be carried out by high pressure liquid chromatograph (HPLC) and in addition, new gels are more powerful than the older Sephadex gels. For example the “Zorbax PSM 1000” silica gel used by Morizur et al. (1984) allowed better recovery of all the organic matter with approximately 1/100 the ion exchange capacity of Sephadex gel. Beckett et al. (1987) used a flow field-flow fractionation (flow FFF) technique for fulvic and humic acid fractionation, which appears to be a powerful tool to obtain information on molecular weights.

418

Organic Analysis

12.3.2 Titration of Functional Groups Section 12.2.2 described titration techniques for the main functional groups of humic compounds, carboxylic and phenolic groups, the latter being obtained by calculating the difference between total acidity and carboxylic acidity. Chemical titration can be used for other functional groups. The main groups that have been the subject of such investigations are: – Total OH groups can be determined by acetylation of humic substances with acetic anhydride in pyridine (Schnitzer and Skinner, 1965): CH3COOR + CH3COOH 2 ROH + (CH3CO)2O The acetylated humic substances are then carefully isolated from the reactional medium. Hydrolysis in alkaline medium releases acetates from the acetyl groups; distillation in strong acid medium enables recovery of acetic acid which is titrated by acidimetry. The number of moles of acetic acid collected corresponds to the number of total OH groups of the humic substance. – Alcoholic OH groups can be estimated by “total OH groups” minus “phenolic OH groups”. However, as the phenolic OH groups are themselves calculated by difference (cf. Sect. 12.2.2), it is advisable to take the laws of propagation of errors into account. – Total C–O Groups can be quantified by the reaction of humic substances with hydroxylamine chlorhydrate in 2-dimethylaminoethanol medium: R1R2C–NOH + H2O + HCl R1R2C–O + NH2OH⋅HCl Excess hydroxylamine hydrochlorate is back titrated with a standard HClO4 solution (Fritz et al. 1959). – Quinonic C–O groups can be titrated by reduction in OH–phenolic groups by ferrous iron in triethanolamine; the excess ferrous iron is back titrated by amperometry with a dichromate solution (Glebko et al. 1970). – Ketonic C–O groups can be estimated by the difference between total C–O groups and quinonic C–O groups. In addition, Schnitzer (1978) cited several attempts to characterize acid groups by direct potentiometric titration. It was difficult to clearly distinguish the two main types of acidity (OH and CO2H functional groups) using this method even in a non-aqueous medium. Certain authors did succeed including Rosell et al. (1972), who simultaneously measured three types of acidity by potentiometric titration in 80% N-methyl acetamid aqueous solution.

Characterization of Humic Compounds

419

De Nobili et al. (1990) presented an alternative to the technique described in Sect. 12.2.2 for the determination of carboxylic groups: the precipitation of humic substances with cethyltrimethylammonium cation detergent. 12.3.3 Characterization by Fragmentation One of the ways to study heavy humic macromolecules consists in splitting them up and identifying the fragments. These methods can be classified in four main groups: oxidative fragmentation, reducing frag mentation, other chemical degradation techniques and thermal fragmenta tion by pyrolysis. Oxidative Fragmentation This group of techniques can be divided into two subgroups. (i) oxidation with permanganate and (ii) other oxidative techniques. Oxidation with permanganate was widely used on humic materials from different types of soils: e.g. Ah horizons of Solonetz, Solod and Chernozem (Kahn and Schnitzer, 1971a; Kahn and Schnitzer, 1972b), forest grey soil under different farming systems (Kahn and Schnitzer, 1972b), tropical volcanic soils (Griffith and Schnitzer, 1975), mediterranean soils (Chen et al. 1978a), non-hydrolysable humic residues (Ogner, 1973) and fulvic acid fractions (Khan and Schnitzer, 1971b). The analytical procedure varies with the author. The humic materials can first be fractionated or subjected to a derivatisation reaction (methylation) before oxidation. Kahn and Schnitzer (1971b) oxidated 1 g of humic material by boiling at reflux for 8h with 250 mL of 4% KMnO4 aqueous solution. The excess of permanganate must be destroyed by controlled addition of small volumes of methanol and the solution removed from insoluble MnO2 by filtration and rinsing. The acidified filtrate is then extracted with ethyl acetate in a liquid-liquid extractor for 48h. The extract is brought to dry in the rotary evaporator, dissolved in a small volume of methanol and methylated with a diazomethane solution in ether. The end products are then split by chromatography and identified with the usual range of spectrographic methods used for molecular characterization (UV, IR, mass and NMR spectrometry), the most widely used method being gas chromatography coupled with mass spectrometry (GC–MS).

420

Organic Analysis

Other oxidative reagents have been used for studies of the degradation of organic matter. Neyroud and Schnitzer (1974) then Griffith and Schnitzer (1976) studied the products of alkaline oxidation of humic and fulvic acids by cupric oxide. Oxidation is performed in an autoclave at 170°C on 1g of humic product mixed with 100 mL of NaOH 2 mol L–1 solution and 5g CuO; the end of the procedure is almost the same as for permanganate oxidation (see earlier part in this section). Other methods included nitrobenzene alkaline oxidation (Morrison 1963), nitric acid oxidation (Hansen and Schnitzer, 1967), hypohalogenite oxidation (Chakrabartty et al. 1974) and peracetic acid oxidation (Schnitzer and Skinner, 1974). Griffith and Schnitzer (1989) reviewed oxidative degradation techniques, one of the analytical tools for the study of humic substances (Hatcher et al. 2001). Reductive Fragmentation The most commonly used reagent was sodium amalgam (Mendez and Stevenson 1966; Stevenson and Mendez 1967; Piper and Posner 1972) but other reducers were also tested such as zinc powder (Hansen and Schnitzer 1969). Stevenson (1989) reviewed reductive fragmentation techniques. Other Degradative Chemical Methods Boiling humic acids in water releases polysaccharides and small quantities of phenolic acids and aldehydes, polypeptides, alkanes and fatty acids (Neyroud and Schnitzer 1975). Acid hydrolysis at reflux boiling enables between 1/3 and 1/2 of organic matter to be dissolved in most soils (Schnitzer, 1978) but this technique was mostly widely used for the study of the organic forms of nitrogen (cf. Sect. 14.2.1 of Chap. 14); Anderson et al. (1978) studied ether-soluble products of acid hydrolysis of fulvic and humic acids. Alkaline hydrolysis was also used as a degradation method for the study of humic molecules. Neyroud and Schnitzer (1975) subjected humic materials to four successive series of hydrolysis with a–NaOH 2N solution in an autoclave at 170°C for 3 h. The recovery and identification of the hydrolysed products was accomplished using techniques similar to those of the other degradation methods (cf “Oxidative Fragmentation”). Parsons (1989) analysed the main fragmentation mechanisms in the acid and alkaline hydrolysis of organic materials. A method developed for depolymerization of coals (Ouchi and Brooks 1967) was applied to degradation of humic acids (Jackson et al. 1972):

Characterization of Humic Compounds

421

reaction with phenol in the presence of p-toluenesulfonic acid as catalyst. A review of degradation techniques by phenol and sodium sulphide was published by Hayes and O’Callaghan (1989). Cheshire et al. (1968) studied the effect on humic acids of alkaline fusion after acid boiling. Thermal Degradation Thermogravimetry (TG), differential thermogravimetry (DTG), differential thermal analysis (DTA) and isothermal heating were used to explore the mechanism of thermal decomposition of humic materials (Schnitzer 1978). Schnitzer and Hoffmann (1964) studied the chemical evolution of humic and fulvic acids under the action of temperatures up to 540°C; Fig. 12.9 shows the curves of differential thermogravimetry these authors observed on their samples.

Fig. 12.9. Curves of differential thermogravimetry (Schnitzer and Hoffman 1964) of organic materials of a podzol: a: surface O2 horizon, b: deep Bh horizon

Kodama and Schnitzer (1970) used differential thermal analysis in a study of the mechanism of thermal decomposition of fulvic acids. Chen et al. (1978b) also used this technique to compare the physicochemical characteristics of humic and fulvic acids of Mediterranean soils (Fig. 12.10).

422

Organic Analysis

Fig. 12.10. Differential Thermal Analysis applied to humic (on the left) and fulvic (on the right) acids extracted from Mediterranean soils (Chen et al. 1978): (a) clayey brown soil, (b) sandy brown soil, (c) clayey silty sandy red soil, (d) sandy silty red soil, (e) silty sandy red soil, (f) sandy silty brown soil

Characterization of Humic Compounds

423

Fig. 12.11. Chromatogram of pyrolysis products of a fulvic acid extracted with soda in a deep Bh horizon of a podzol (Martin, 1976)

Fig. 12.12. Mass spectra of pyrolysis obtained by Saiz-Jimenez et al. (1979) on a sample of brown soil on granite rock (typical xerochrept); the humin residue fraction was the result of alkaline extraction of the soil following humic acid extraction and complete destruction of silicates by the HF– HCl mixture

424

Organic Analysis

Pyrolysis–gas chromatography was developed by Kimber and Searle (1970) for the study of soil organic matter. Figure 12.11 shows one of the chromatograms of pyrolysis obtained by Martin (1976) on fulvic acids of a deep Bh horizon of podzol. The pyrolysis mass–spectrometry described by Meuzelaar et al. (1973) was applied to humic compounds by several authors (e.g. Meuzelaar et al. 1977 Saiz-Jimenez et al. 1979). The humic sample was dispersed in a soda solution or in methanol, covered and placed on a ferromagnetic coil and pyrolyzed at 510°C. An example of the mass spectrum of the pyrolysis products is shown in Fig. 12.12. Reviews by Bracewell et al. (1989) and Schulten (1996) give more data on the thermal degradation products of humic materials. 12.3.4 Nuclear Magnetic Resonance (NMR) Principle The majority of atomic nuclei turn around their axis and thus have an angular moment expressed by the formula [h/2π][I(I+1)]1/2 where h is the Planck constant and I the spin number. The spin number values can be 0, 1/2, 1, 3/2, 2, etc., depending on the nature of the nucleus: for 1H and 13C, I = 1/2, but for 12C, I=0. Because of the electric charge, the rotation of the nucleus creates a magnetic field. Conversely, if a nucleus is placed in a magnetic field Ho it can orient itself in one of the (2 I +1) directions linked to the direction of the field. Each direction corresponds to an energy state and it is possible to induce a resonance between the energy states by using electromagnetic radiation of a frequency ν such as:

ν = γ Ηο / 2π,

(12.1)

where the gyromagnetic ratio γ is a constant that depends on the type of nucleus.
Table 12.1 Relative detectability (RD) of a nucleus in soil organic matter by NMR (Wilson 1981)

Nucleus 1 a H
27 23

RD 102 101 100

Al

Na

Characterization of Humic Compounds
1

425

Hb Mn Si C N K O

55

100 10-1 10-1 10-3 10-3 10-3 10-3 10-3 10-3 10-4 10-4 10-5 10-6

29

13 14 39 17 25

Mg Zn P Ca Fe N

67

31 43

57

15 a

: very variable, depends on the water contents and the pH
b

: H of soil organic matter only

The atom of hydrogen (I = 1/2) gives (2I + 1) = 2 possible orientations of the nucleus and it is possible to detect its resonance. On the other hand, most nuclei are not detectable: for 12C, (2I + 1) = 1 thus there is only one possible orientation of the nucleus in the magnetic field and it is not possible to induce a resonance. Only the 13C isotope of carbon can be studied but is much less abundant than the 12C isotope in the natural state. Finally, the most important factors in measuring the detectability of an element in soil or in soil extracts are the spin number I, and the gyromagnetic ratio γ but also the abundance of the element (n) and the relative abundance of the isotope under study (N). Table 12.1 presents the detectability of atomic nuclei in soil as a function of a calculation by Wilson (1981) with the formula: γ 3NI(I+1)n. This table does not give the detectability of an element itself but its detectability in the soil or in a soil organic matter medium. Thus, an element that is sensitive to NMR such as 31P will be detected with difficulty because of its low concentration.

426

Organic Analysis

The traditional NMR approach consists in subjecting the sample to a radio frequency scan (continuous wave NMR), with a fixed magnetic field (or vice versa) and recording resonance when the irradiation frequency matches the frequency of nuclear transition given by (12.1). The Fourier Transform NMR often enables high quality spectra to be obtained more rapidly. In FTNMR, the nucleus is subjected to short and intense pulsation of radiation and its behaviour is observed. All the nuclei resound simultaneously and the resulting spectrum of the signal as a function of time (free induction decay, FID) is not very useful for the chemist. Signal processing by Fourier transformation should be used to obtain more easily interpretable spectra of the continuous wave NMR type. To increase sensitivity, a large number of “FID” has first to be collected on the computer and then an average used to calculate the Fourier transform. The NMR technique would provide little useful information for structural chemistry if only the spin transition(s) of the nucleus were measured at frequencies corresponding to (12.1). In fact, the electronic environment of the nucleus protects it from the applied magnetic field (Ho) and the real magnetic field at constant frequency (or the frequency at constant field) necessary for nuclear resonance depends on how effectively the nucleus is protected. In organic molecules, the functional groups have different electronic distributions and the frequencies of resonance of each of their nuclei shift slightly depending on the nature of the functional group. It is then possible to identify these groups. In practice, the shift in frequency is identified by comparing it to a standard, usually tetramethylsilane (TMS), in terms of chemical shift (δ ) calculated by the ratio of the “chemical shift of frequency” compared to the “frequency of the standard”. As the chemical shifts are weak, δ are generally noted in parts per million: δ(ppm). Another significant parameter is relaxation. After the excitation of a nucleus, some time passes before it returns in its fundamental energy state. In theory, this return occurs in two ways: by interaction with the molecular lattice or by spin energy exchange with “brother” nuclei. The time constants corresponding to these two processes are named spinlattice relaxation time and spin–spin relaxation time (T1 and T2, respectively). This phenomenon is significant for the soil chemist. In 13C NMR studies, it can affect the quantitative measurement of the carbon of the functional groups.

Characterization of Humic Compounds

427

Following the development of this technique NMR was used for the study of soil organic matter. Originally extracts were analysed in solution by studying 1H proton NMR then 13C NMR, first qualitatively, and then while trying to quantify the observations. The material had to be soluble in a solvent which did not give resonance to the element under study (e.g. for studies of 1H, D2O was used instead of H2O), but this is no longer indispensable (Wilson, 1981). More recent studies use NMR on the solid phase. Wilson (1981, 1987) reviewed these methods and their use for the study of soil organic matter. Steelink et al. (1989) provided complementary information in the field of 1H and 13C NMR of humic substances in solution. Wilson (1989) and Tate (1998) provided additional theoretical information in the field of solid state NMR and its use for humic substances. Simpson (2004) applied coupling NMR and separation techniques. 15N NMR was applied to the study of the nitrogen cycle (Thorn and Mikita 2000). Study of Humic Materials by 1H NMR Schnitzer and Barton (1963) were the first to observe NMR spectra of organic extracts of soil. They used spectrometry with continuous wave scanning applied to fractions of methylated fulvic acids which provided relatively poor structural information. Subsequently, a relatively large number of authors used the same wave scanning technique 1H NMR. Lentz et al. (1977) obtained spectra of better quality with the Fourier transform technique. Wilson et al. (1978) used the techniques with Fourier transform in a high magnetic field at a very high frequency thereby further increasing the information provided by 1H NMR spectra (Fig. 12.13). Study of Humic Materials in Solution by 13C NMR C NMR has several advantages over 1H NMR for the structural analysis of molecules: it provides direct information on the structural skeleton, which enables observation of functional groups without protons like ketones. The carbon nucleus provides more significant chemical shifts enabling detection of finer differences in molecular structures. The signals can also include narrower peaks, and this reduces the risk of one peak concealing another. On the other hand, the disadvantage of 13C NMR lies in the small proportion of 13C isotopes (only 1.1% of carbon) leading to difficulty in detecting the signals (Wilson 1981).
13

428

Organic Analysis

For satisfactory detection, the conditions required to obtain the 13C spectra always have to be optimized. These spectra are almost always obtained under conditions of decoupling with protons (Proton Decoupled 13 C NMR) which induces exaltation of 13C resonance (Nuclear Overhauser Enhancement NOE). The NOE and dipolar 13C–1H relaxation theory in the proton decoupled 13C NMR spectra of the macromolecules was studied by Doddrell et al. (1971).

Fig. 12.13. Fourier transform 1H NMR spectrum obtained at 270 MHz by Wilson –1 et al. (1978) on humic materials extracted by a 0.1 mol (NaOH) L solution on a silty potter’s clay from Wakanui. A pulsation of 10 µs with an inclination angle of 85° was used to visualize a spectral window of 3.6 kHz in 8 ko of data points.

The use of these techniques for soil organic materials was initially unsuccessful (Schnitzer and Neyroud 1974). Subsequently several authors tried to improve the technique, and Newman et al. (1980) achieved noticeable improvement in the quality of the spectra (Fig. 12.14) by optimizing the operating conditions, particularly the spacing between the radiation pulsations. The optimum interval of 0.2 s also corresponded to the acquisition time of the relaxation signals coming from 2 105 90° impulses of 23 µS. The spectra were obtained on an Varian FT-80A apparatus operating at 20 MHz for 13C with proton decoupling at 80 MHz. Acquisition required collection of 4,000 (coal samples) to 136,000 (soil humic acid samples) free induction decay spectra then multiplication by a filter with a 20 ms time constant, before the Fourier transform. The samples were prepared for analysis by suspension of 300 mg of humic

Characterization of Humic Compounds

429

acid in 2 cm3 of 0.5 mol (NaOH) L–1 solution for 1 day at 20°C followed by ultracentrifugation (84,500 g at 4°C). After dilution to 50% with D2O, RM–13C was measured in tubes with a diameter of 10 mm.

Fig. 12.14. Proton decoupled 13 C NMR spectrum of a humic acid in solution ob tained by Newman et al. (1980) with optimized acquisition parameters (see conditions described in the text)

Newman and Tate (1984) used very similar conditions to those described above to characterize the humic substances of alkaline extracts of soils, the total time needed for spectral acquisition ranged from 10 to 50 h. Preston and Schnitzer (1984) studied the effect of the type of extraction (acid or alkaline extraction) and of chemical modifications in the extracted material (methylation, hydrolysis using 6 mol (HCl) L–1 followed or not by methylation) on 13C NMR spectra of humic materials from four types of soils. The materials were dissolved in deuterated solvents (CDCl3 for methylated materials, 0.5 mol (NaOD) L–1 (D2O) solution for non-methylated materials). The chemical shifts were measured compared to the sodium 3-trimethylsilylpropionate (TSP) for the heavy water (D2O) solutions and with tetramethylsilane (TMS) for the deuterated chloroform (CDCl3) solutions. The spectra were obtained on a Bruker WM 250 spectrometer on 10 mm sample tubes with an interruption technique of 1H decoupling except during the acquisition time (inverse-gated decoupling of Freeman et al. 1972). Some of spectra were comparable with those shown in Figure 12.14.

430

Organic Analysis

Study by Solid State 13C NMR The extraction of organic matter (cf. Sect. 11.2.1 and 11.3.1 of Chap. 11) can result in molecular modifications, particularly if strong bases are used as extraction reagents. Moreover, most materials remain in a nonextractable state in the humin residue. The in situ study of organic matter allows these disadvantages to be overcome. However, conventional NMR spectroscopy applied to whole soil resulted in only broad and diffuse signals: many dipole–dipole interactions produced signals comprising information on the chemical shifts that was not clear, and moreover, the spin–lattice relaxation times (T1) needed too long to accumulate the free induction decay signals required to obtain quantitative information. The theoretical advance in NMR technique (Pines et al. 1973) subsequently made it possible to overcome the problem and envisage new developments. In the cross polarization or CP13 C NMR technique, the protons are uncoupled from the 13C nucleus and used to increase the relaxation kinetics of the 13C nucleus. The signal peak widths can be reduced to a degree where the functional groups can be partially identified. However, the detection of the carbon forms in the whole soil using the CP-13C NMR method required very organic soils (6% of C according to Wilson 1981). This limit has gradually been reduced thanks to improvements in the quality of the instruments and also the preliminary use of physical fractionation methods such as those described in Chap. 9, spectral analysis being limited to the most organic fractions (Barron et al. 1980). The magic angle spinning (MAS) NMR technique, which was originally described by Lowe (1959) and subsequently applied to polymers by Schaefer and Stejskal (1976), can also help get round problems like spectral resolution, and obtain spectra directly on the soils; in this technique, the sample is put in fast rotation at an axial slope of 54°44¢, which means anisotropic effects can be reduced and isotropic shifts can be selected. This technique is often used together with cross polarization, and is then known as the CPMAS–13C NMR method. For soil studies, most authors preferred to apply NMR techniques to previously concentrated humic materials, usually humic or fulvic acids. A study by Newman et al. (1980) clearly showed the difference in resolution that still existed between the 13C NMR techniques in solution and CP 13C NMR on the solid phase (Fig. 12.15a).

Characterization of Humic Compounds

431

Thanks to improvements in this technique, the spectra of CPMAS–13C NMR obtained by Gerasimowicz and Byler (1985) on humic substances showed a better resolution (Fig. 12.15b) than that of Newman et al. (1980). Fründ and Lüdemann (1989) performed instructive comparisons between the technique in solution and CPMAS–13C NMR which showed that the degree of detail provided by the second technique was similar to that of the first; in addition, the spectra obtained on a rendzina soil (4.6% of carbon) and on its humic extracts and humin residue were compared under satisfactory conditions (Fig. 12.15c). The two techniques (liquid phase NMR and solid phase CPMAS–13C NMR) were recommended by Conte et al. (1997a) for the study of organic materials of the soils. State-of-theart CPMAS–13C NMR allows observation of organic materials in their environment (without fractionation) when the soils are not too low in carbon, as attempted by Kinchesh et al. (1995) on Rothamsted soils (UK). Other studies such as that of Conte et al. (1997b) on volcanic soils used CPMAS–13 C NMR on extracted humic substances. 3.4.5 Quantification of Observations by NMR Different techniques exist for the quantification of the information provided by the NMR signals of humic substances. The oldest derive from the study of coal and coal-like materials. The method of Brown and Ladner (1960) enables estimation of the aromaticity of these carbonaceous materials using the 1H NMR spectrum. Wilson (1981) proposed an adaptation of this method for use on humic materials. The rate of aromaticity fa was most often studied by quantification of 13 C NMR signals. However, techniques for the direct quantification of the different peaks of the spectra should be used with caution. The 13C atoms of the different functional groups have different nuclei relaxation times, the nuclei with the shortest times making a weaker contribution to the total spectrum than those with the longest time. In addition, due to proton decoupling, NOE results in an increase in the signal that differs with the nature of the nucleus (Wilson 1981). However, these two factors were sometimes shown to have no significant influence on the direct measurement of the rate of aromaticity (Newman et al. 1980).

432

Organic Analysis

Fig. 12.15 Improvements in solid phase 13C NMR techniques: (a) CP 13C NMR . spectra obtained by Newman et al. (1980) on solid coal, a solid humic 13 acid (HA), and a solution of H A, (b) CP MAS- C NMR spectra obtained by Gerasimowicz and Byler(1985) on HA from muds at different stages of composting treatments (C) CPMAS-13C NMR spectra obtained by Früd and Lüdem ann (1989) on a rendzina soil, n on its fulvic (FA) and humic (HA) extracts and on the humin residue

Characterization of Humic Compounds

433

Fründ and Lüdeman (1989) improved quantitative analysis of humic materials. Their method enabled simultaneous measurements of carboxylic-, aromatic-, carbohydrate- and aliphatic-C rates. Smernik and Oades (2000a) highlighted the effect of paramagnetic impurities and of purifications by HF treatment (Smernik and Oades, 2000b). Smernik and Oades (2003) then Moser and Lefebvre (2004) explored new ways to improve 13C NMR quantification on soil organic matter. Conte et al. (2004) reviewed state-of-the-art CPMAS C-13-NMR spectroscopy applied to natural organic matter. 12.3.5 Fluorescence Spectroscopy Although less widely used than the visible UV (cf. Sect. 12.2.3) or infrared (cf Sect. 12.2.4) absorption spectrometry, fluorescence spectrometry has been tested by several authors as a complementary technique to characterize humic substances. Lévesque (1972) used this technique on Fe and P humic complexes. The emission spectrum of a fulvic acid has a main peak with a not very variable wavelength that moves from 500 to 520 nm when the excitation wavelength moves from 400 to 468 nm. As emission spectra generally provide little information, excitation spectra were used instead. The emission wavelength was then fixed at around 500–520 nm and excitation varied in a continuous way between 250 and 500 nm. Ghosh and Schnitzer (1980a) observed on their humic substances two quite distinct excitation bands, at 465 and 360 nm, and a decrease in intensity of the fluorescence with an increase in the ionic force and a reduction in the pH; their data were subsequently used to calculate the constant of dissociation of humic acids (Goldberg et al. 1987). Bachelier (1981) refined these observations with a precise description of the excitation spectra on a larger number of soils; he noted the presence of seven peaks (including two rare ones) corresponding to slope changes on the large 465 nm peak and of two peaks (including one rare one) on the large 360 nm peak. Bachelier also studied the fluorescence of several types of humic acids split on G25 Sephadex gel (1) humic acids of high molecular weight which are eluted at the head of the column (coarse humic acid cHA) give a weak fluorescent signal; (2) the following band of less condensed brown-yellow humic acids gives bright yellow fluorescence under UV radiation (fluorescent humic acids f HA); (3) lower fluorescence compounds called higher humic acids (hHA) are sometimes found after this band and (4) intermediate humic acids (iHA) are sometimes found

434

Organic Analysis

between the two cHA and fHA bands. Figure 12.16 shows some excitation spectra obtained on different humic products. Bachelier’s observations confirm those of Lévesque on the low fluorescence of the humic complexes with high molecular weight.

Fig. 12.16. Fluorescence excitation spectra of humic materials observed in the 509–515 nm emission band (Bachelier, 1981): (FA) fulvic acid of various samples of soils, (HA) different fractions of humic acids separated by gel chromatography (see fractionation diagram above the graph)

Characterization of Humic Compounds

435

Bachelier (1984) used fluorescence spectrometry to distinguish the degrees of condensation of humic acids. Nayak et al. (1985) studied the fluorescence of humic acids in different solvents. The fluorescence polarization technique enabled further information to be obtained on the aggregation and conformation of humic molecules and was used to this end in a study of fulvic acid by Lapen and Seitz (1982). 12.3.6 Electron Spin Resonance (ESR) Spectroscopy When molecules containing unpaired electrons are placed in a magnetic field, interaction occurs between the magnetic moment of these electrons and the applied field which results in decoupling into two discrete states of energy of each unpaired electron. This is electron spin resonance (ESR) sometimes still called electron paramagnetic resonance (EPR). ESR spectroscopy uses electromagnetic excitation radiation located in the spectral field of microwaves; it is used for the study of compounds containing unpaired electrons, primarily free radicals. The energy of an unpaired electron in a magnetic field is given by E = –g β Mz Ho, where G represents the spectroscopic split factor which has a value of 2.0023 for a free electron, β the magneton of Bohr, Mz the component of the angular spin moment in the direction of the z-axis of the applied magnetic field which can take the discrete values +1/2 and –1/2, Ho the force of the magnetic field. For a given value of Ho the difference in energy ∆E between two states of discrete spin of the electron is ∆E = g β Ho the phenomenon of resonance occurs when this energy is equal to that of the applied field: ∆E = hν and H being Planck’s constant and ν the frequency. In his pioneer work, Rex (1960) used ESR spectroscopy to highlight the radical nature of humic molecules. MacCarthy and Rice (1985) listed several works reporting the use of ESR spectroscopy for humic substances. The spectra of humic substances are generally simple signals identified by their position and width (Fig. 12.17). The hyperfine lines that can be identified in some molecules are not usually present in humic spectra.

436

Organic Analysis

Fig. 12.17 Electron spin resonance spectrum of a fulv acid (Schnitzer and . ic Skinner, 1969)

Schnitzer and Skinner (1969) showed on ten humic materials that the most variable parameter deduced from these spectra is the spin concentration. This is determined by comparison with a standard of calibration, the number of radicals being proportional to the signal surface. In humic substances, the concentration is found around 1018 spins g–1. Schnitzer and Skinner also tested the influence of several factors on the spin concentrations of humic substances (chemical modifications, heating) as well as the relation between other parameters and the spin concentration (molecular weight, E4:E6 ratio, number of molecules of a given weight per free radical). Riffaldi and Schnitzer (1972) studied the effect of the experimental conditions on the ESR spectra of humic substances to highlight mistakes due to too rapid interpretation of these spectra. MacCarty and Rice (1985) also commented on the relative poverty of the information from ESR spectrometry for the study from the humic substance. However, subsequent technological developments changed this situation. For example, Senesi et al. (1989) provided more detailed spectra than previous authors and Senesi (1990) reviewed stateof-the-art and potential development of ESR technique in its application to soil chemistry. Saab and Martin-Neto (2004) studied semiquinone free radicals of humic substances by ESR.

Characterization of Humic Compounds

437

12.3.7 Measurements of molecular weight and molecular size Principles Measurement of molecular weight is a traditional procedure in structural chemistry. As a complement to ultimate analysis, the molecular weight provides an empirical formula for the compound concerned before the determination of the structural formula (functional groups) using spectroscopic and chemical methods. The measurement of molecular weight and size has been attempted many times on humic substances and several reviews have been devoted to the subject (e.g. Orlov et al. 1975; Stevenson 1982; Wershaw and Aiken 1985; Buurman 2001). It has also been used to measure the humic substances of water (Yu-Ping Chin et al. 1994, Yamada et al. 2000) or atmosphere (Samburova et al. 2005) However, the results differed considerably firstly depending on the technique used, and secondly because, as emphasised by the authors of these reviews, humic substances are not discrete chemical entities but complex mixtures of polydisperse organic substances with a wide range of molecular size. The problem with measuring the molecular weight of mixtures nevertheless was studied during many years. In 1935 Lansing and Kraemer pointed out that the most commonly used methods resulted in average molecular weights and that, depending on the method of measurement used, these average weights could not always be compared. Three types of average molecular weights corresponding to three types of measurement are now used: Number-Average Molecular Weight Mn This is obtained using methods that measure the number of molecules (generally in a diluted solution) irrespective of their size. It is expressed by: Mn = Σni Mi / Σ ni, where ni is the number of molecules of molecular weight Mi Mn is determined by all measurements corresponding to a thermodynamic property connected to a number of molecules in solution (colligative

438

Organic Analysis

property): lowering of vapour pressure, lowering of the freezing point, rise in the boiling point (Raoult laws), osmotic pressure. Weight-Average Molecular Weight Mw Weight-average molecular weight is measured by methods concerned with the masses of different materials, like light scattering and ultracentrifuge sedimentation; it is expressed by: Mw = Σ wi Mi / Σ wi = Σ ni Mi2/Σ ni Mi, where wi represents the mass fraction of each species; The z-Average Molecular Weight Mz This can also be calculated with the ultracentrifugation data obtained and is expressed by: Mz = Σ wi Mi2 / Σ wi Mi = Σ ni Mi3/Σ ni Mi2. In a monodisperse system Mn = Mw = Mz but this is not the case in polydisperse systems; the number-average molecular weight then tends to represent the lowest molecular weights whereas Mw tends to represent the heaviest particles of the mixture (Wershaw and Aiken, 1985). Orlov et al. (1975) expected Mw to be better correlated with the known properties of humic substances. In heterogeneous systems, Mz > Mw > Mn, the Mw:Mn ratio can generally be used to calculate the degree of polydispersity (Stevenson 1982). In addition to methods for the measurement of the average molecular weight, another group of methods measures the size of the molecules, e.g. gel filtration, ultrafiltration and small angle X-ray scattering. In these methods, model compounds of known molecular weight and composition are used to determine the molecular weight of humic substances, but problems can occur if these compounds are too different from the humic molecules (Wershaw and Aiken 1985). Methods for the Measurement of Molecular Size Gel Exclusion Chromatography The methods of gel exclusion (or gel permeation) chromatography are described in Sect. 12.2.1 and 12.3.1. It should be noted that Reuter and Perdue (1981) found a very big difference between the expected molecular weights on humic fractions of Sephadex gel and the numberaverage molecular weights actually measured. Plechanov (1983) used the

Characterization of Humic Compounds

439

Sephadex LH60/dimethylformamide/acetic acid system for the measurement of the molecular weight of humic substances. Nobili et al. (1989) reviewed the use of gel chromatography for the measurement of the molecular size of humic substances. Ultrafiltration Ultrafiltration by pressure filtration through a membrane is also used for the separation of macromolecules as a function of their molecular size. This technique is similar to reverse osmosis, except with respect to the size of the particles that can be split: reverse osmosis separates the particles of molecular size near to those of the solvent whereas ultrafiltration separates particles approximately 10 times the size of the solvent i.e. up to 0.5 µm (Wershaw and Aiken 1985). A large number of different types of membranes exist which are classified by their manufacturers according to their threshold of cut expressed in molecular weight. Nevertheless ultrafiltration is not a technique for separation by weight but by molecular size. A review by Wershaw and Aiken (1985) provided a lot of useful information about this technique. Scattering of Electromagnetic Radiation The principle of this technique for light scattering was described by Kerker and Milton (1968) and by Guinier and Fournet (1955) for small angle X-ray scattering. The reader is advised to consult these publications for a comprehensive explanation of the phenomenon and to refer to Wershaw and Aiken (1985) and Wershaw (1989) for the use of this technique for humic substances. Methods of Measurement of Molecular Weight Determination of the Number-Average Molecular Weight by Measurement of Colligative Properties By definition, a colligative property is a thermodynamic property which depends on the number of particles in a solution independent of their nature. At the infinite dilution limit, each one of these properties is proportional to the number of molecules of solute present in the solution. The classical theory of each colligative property is described found in all chemistry and physics handbooks. The most widely used techniques for the measurement of molecular weight of humic substances are cryoscopy and vapour pressure osmometry. Cryoscopy records the drop in the temperature during solidification of the solvent in the presence of the solute to be studied. Vapour pressure osmometry measures the change in osmotic pressure resulting from the passage of a solvent through a

440

Organic Analysis

membrane from a diluted solution to more concentrated one. For a description of these methods and their use for humic substances, see Stevenson (1982), Wershaw and Aiken (1985) and Aiken and Gillam (1989). Ultracentrifugation There are two distinct groups of ultracentrifugation techniques, those concerned with sedimentation kinetics, and those concerned with ultracentrifugation equilibrium. The first group was the most commonly used on humic substances (Wershaw and Aiken 1985) sometimes as a complement to other fractionation methods (Cameron et al. 1972b), although the centrifugation equilibrium provides a wealth of information since Mn, Mw and Mz can be determined at the same time (Posner and Creeth 1972). The theory and implementation of the first technique is detailed in Cameron et al. (1972b), the second by Posner and Creeth (1972), and a review by Swift (1989) gives a detailed description of all ultracentrifugation techniques. Viscosimetry Measurements of viscosity can provide significant information concerning the size and the shape of the molecules. The well-known Oswald viscometer records the times of passage of the solution and solvent between two reference points marked on the apparatus. The molecular weight can be estimated from measurements of viscosity using the Staudinger equation: [η] = kMα, in which intrinsic viscosity η is linked to molecular weight M by two adjustable parameters (Ghosh and Schnitzer, 1980b). Techniques based on measurement of viscosity were reviewed by Clapp et al. (1989). Flow FFF (cf. Sect. 12.3.1) was also once considered to be promising for the measurement of molecular weight of humic substances (Beckett et al. 1987). 12.3.8 Microscopic Observations Several studies report observations of humic materials using optical microscopy, transmission electron microscopy and scanning electron microscopy. The first difficulty is linked with the different ways the sample can be modified depending on the preparation techniques (e.g. the degree of separation of inorganic materials, molecular modifications with respect to the ionic force and pH), and the second difficulty is the conditions of the observation itself (heating of the sample). Bachelier (1983) carried out observations on nine different types of soil using three

Characterization of Humic Compounds

441

techniques: frozen humic acid solutions under a binocular magnifying glass, desiccated humic acid solutions under a transmission electron microscope, humic acid solutions freeze-dried on gilt aluminium film under a scanning electron microscope. The latter requires metallization of the non-conducting substances before observation. Chen and Schnitzer (1976) studied the influence of pH on the appearance of humic acids by scanning electron microscopy while Stevenson and Schnitzer (1982) studied the same effect by transmission electron microscopy. Chen and Schnitzer (1976) used transmission electron microscopy for the study of metallic complexes of fulvic acids. Scanning electron microscopy was used by Chen et al. (1978) for the comparison of humic acids from Mediterranean soils. Tan (1985) provided detailed methodological information particularly concerning the preparation of the samples. 12.3.9 Other Techniques The main techniques used for structural characterization are described in this chapter. However other techniques can also be used to improve our knowledge of the structures of humic compounds and their linkage with mineral materials. X-ray techniques are not limited to small angle X-ray scattering described in “Scattering of Electromagnetic Radiation” in Sect. 12.3.7; Xray diffraction was also used by Schnitzer (1978), and can provide useful information in spite of the non-crystalline character of humic substances. As far as electrochemical methods are concerned, a study of humic acid characterization by polarography was described by Shinozuka and Hayano (1987). The use of FTIR spectroscopy makes it possible to increase the resolution and to decrease the background noise of the IR spectra. However, these advantages may not be apparent in the study of humic substances because of their molecular nature (MacCarthy and Rice, 1985). More recent techniques like X-ray photoelectron spectroscopy (XPS) and Mössbauer spectroscopy have not been widely used for the study of humic substances. XPS, also called electron spectroscopy for chemical analysis (ESCA), can only be used with solid materials because it requires a high vacuum; it is based on the analysis of the electrons emitted by the internal electron-shells atoms when they are subjected to X-ray bombardment of sufficient energy; Defosse and Rouxhet (1980) introduced this technique in soil analysis.

442

Organic Analysis

Mössbauer spectroscopy is not directly useable for the study of the structure of metal compounds, but can be used for the study of organometallic bonds, and particularly for the study of complexes of iron–humic compounds (Goodman and Cheshire, 1979). It will be interesting to see whether future progress results from new technologies or from synthetic studies on molecular models as was the case for the identification of the double helix of ADN by Watson and Creek.

References
Molecular Models
Hatcher PG, Dria KJ, Sunghwan Kim, Frazier SW (2001) Modern analytical studies of humic substances. Soil Science, 166, 770–794 Lawson GJ and Stewart D (1989) Coal Humic Acids. In Humic substances II, search of structure, Hayes HB., MacCarthy P, Malcolm RL and Swift RS ed., Wiley, 641–686 Lodygin ED and Beznosikov VA (2003) The 13C NMR study of the molecular structure of humus acids from podzolic and bog-podzolic soils. Pochvovedenie, 9, 1085–1094 Piccolo A and Conte P (2003) Comments on “Modern analytical studies of humic substances” by Hatcher et al. Soil Sci., 168, 73–74 Schulten HR (1995) The three-dimensional structure of humic substances and soil organic matter studied by computational analytical chemistry. Fresenius J. Anal. Chem., 351, 62–73 Schulten HR and Leinweber P (2000) New insights into organic-mineral particles: composition, properties and models of molecular structure. Biol. Fertil. Soils, 30, 399–432 Schulten HR and Schnitzer M (1997) Chemical model structures for soil organic matter and soils. Soil Sci., 162, 115–129 Skjemstad JO, Clarke P, Taylor JA, Oades JM and McClure SG (1996) The chemistry and nature of protected carbon in soil. Aust. J. Soil Res., 34, 251–271 Schulten HR (2002) New approaches to the molecular structure and properties of soil organic matter: humic-, xenobiotic-, biological-, and mineral-bonds. In 3rd Symposium on Soil Mineral–Organic Matter–Microorganism Interactions and Ecosystem Health, Naples-Capri, Italy, 22–26 May 2000, Violante A, Huang PM, Bollag JM and Gianfreda L. ed. Developments in soil science, volume 28A, 351–381

Characterization of Humic Compounds

443

McCarthy P (2001) The principles of humic substances. Soil Sci., 166, 738–751 Piccolo A, Celano G, Conte P, Zena A and Spacini R (1999) Adsorption of atrazine on humic substances of different molecular structure and their hydrolysed products as modified by pH. In Human and environmental exposure to xenobiotics Del Re AAM, Brown CD, Capri E, Evans SP and Trevisan M, ed., Proceedings of the XI Symposium of Pesticide Chemistry, Cremona, Italy, September 12–15 1999, 425–431

Fractionation, Determination of Molecular Weights and Molecular Sizes
Aiken GR and Gillam AH (1989) Determination of molecular weights of humic substances by colligative property measurements. In : Humic substances II., Hayes MHB., MacCarthy P, Malcolm RL and Swift RS ed. Wiley New York, 515–544 Bailly JR and Margulis H (1968) Etude de quelques acides humiques sur gel de dextrane. Plant and soil, XXIX, 343–361 Bailly JR and Tittonel E (1972) Etude de quelques acides humiques sur gel de dextrane (II). Plant Soil, 37, 57–80 Beckett R, Zhang Jue and Giddings JC (1987) Determination of molecular weight distributions of fulvic and humic acids using flow field-flow fractionation. Environ. Sci. Technol., 21, 289–295 Buurman P (2001) Understanding humic substances: advanced methods, properties and applications. Soil Sci., 166, 950–951 Cacco G and Maggioni A (1976) Isoelectrophoretic characterization of humic and fulvic acids extracted from an alpine podzol. Agrochimica, 20, 20–28 Cacco G, Maggioni A and Ferrari G (1974) Electrofocusing : a new method for characterization of soil humic matter. Soil Biol.Biochem., 6, 145–148 Cameron RS, Swift RS, Thornton BK and Posner AM (1972a) Calibration of gel permeation chromatography materials for use with humic acid. J. Soil Sci., 23, 343–349 Cameron RS, Thornton BK, Swift RS and Posner AM (1972b) Molecular weight and shape of humic acid from sedimentation and diffusion measurements on fractionated extracts. J. Soil Sci., 23, –342 Clapp CE, Emerson WW and Olness AE (1989) Sizes and shapes of humic substances by viscosity measurements. In : Humic substances II., Hayes MHB, MacCarthy P, Malcolm RL and Swift RS ed. Wiley New York, 497–514 Dabin B (1980) Les matières organiques dans les sols tropicaux normalement drainés. Cah.ORSTOM sér.Pédol., 18, 197–215 Dabin B and Thomann Ch (1970) Etude comparative de deux méthodes de fractionnement des composés humiques (méthode Tiurin et méthode électrophorétique). ORSTOM ser. Initiation-documentation technique No.16, 66 p

444

Organic Analysis

Duchaufour Ph and Jacquin F (1963) Recherches d’une méthode d’extraction et de fractionnement des composés humiques contrôlés par électrophorèse. Ann. Agron., 19, 6 Duchaufour Ph and Jacquin F (1966) Nouvelles recherches sur l’extraction et le fractionnement des composés humiques. Bull. ENSAN, VIII, 1, 3–24 Duchaufour Ph (1954) Propriétés des complexes humiques dans différents types de sols. Ecole Nationale eaux et forêts, Nancy, 29 p Duchaufour Ph (1956) Pédologie: Applications forestières et agricoles. Ecole Nationale eaux et forêts, Nancy, 310 p Duchaufour Ph (1957) Pédologie: tableaux descriptifs et analytiques des sols. Ecole Nationale eaux et forêts, Nancy, 87 p Duxbury JM (1989) Studies of the molecular size and charge of humic substances by electrophoresis. In : Humic substances II., Hayes MHB, MacCarthy P, Malcolm RL and Swift RS ed. Wiley New York, 593–620 Ghosh K and Schnitzer M (1980b) Macromolecular structures of humic substances. Soil Sci., 129, 266–276 Guinier A and Fournet G (1955) Small angle X-ray scattering., Wiley New York, 268 p Kerker and Milton (1968) Light scattering. Ind. Eng.Chem., 60, 31–46 Lansing WD and Kraemer EO (1935) Molecular weight analysis of mixtures by sedimentation equilibrium in the Svedberg ultracentrifuge. J. Am. Chem. Soc., 57, 1369–1377 Laurent TC and Killander J (1964) A theory of gel filtration and its experimental verification. J. Chromatog., 14, 317–330 Morizur JP, Monegier du Sorbier B, Silly L and Desbene PL (1984) Etude par chromatographie sur gel avec détection spectrométrique de l’évolution de la conformation des acides humiques en fonction de la force ionique et du pH. C. R. Acad. Sc. Paris, 299, 1269–1272 Nobili Maria De, Gjessing E and Sequi P (1989) Sizes and shapes of humic substances by gel chromatography. In : Humic substances II., Hayes MHB, MacCarthy P, Malcolm RL and Swift RS ed. Wiley New York, 561–591 Orlov DS, Ammosova YaM, Glebova GI (1975) Molecular parameters of humic acids. Geoderma, 13, 211–229 Plechanov N (1983) Studies of molecular weight distributions of fulvic and humic acids by gel permeation chromatography. Examination of the solute molecular composition using RI, UV, Fluorescence and weight measurement as detection techniques. Org. Geochem., 5, 143–149 Posner AM and Creeth JM (1972) A study of humic acids by equilibrium ultracentrifugation. J. Soil Sci., 23, 333–341 Ratsimbazafy CA (1973) Protocole de fractionnement et d’étude de la matière organique des sols hydromorphes de Madagascar. Cah. ORSTOM sér. Pédol., XI, 227–236 Reuter JH and Perdue EM (1981) Calculation of molecular weights of humic substances from colligative data: application to aquatic humus and its molecular size fractions. Geochim. Cosmochim. Acta, 45, 2017–2022

Characterization of Humic Compounds

445

Righetti PG and Drysdale JW (1971) Isoelectric focusing in polyacrilamide gels. Biochem. Biophys. Acta., 236, 17–28 Rusina TV, Kasparov SV and Zharikov AV (1983) Method of electrophoretic research of humus substances and proteins in soi l solutions (texte Russe, résumé Anglais). Pocvovedenie, 1, 38–46 Samburova V, Zenobi R and Kalberer M (2005) Characterization of high molecular weight compounds in urban atmospheric particles. Atmos. Chem. Phys., 5, 2163–2170 Stevenson FJ (1982) Colloidal properties of humic substances. In : Humus chemistry, Stevenson FJ ed. Wiley and Sons, 285–308 Swift RS (1989) Molecular weight, shape, and size of humic substances by ultracentrifugation. In : Humic substances II., Hayes MHB, MacCarthy P, Malcolm RL and Swift RS ed. Wiley New York, 467–495 Tiurin (1951) Vers une méthode d’analyse par l’étude comparative des constituants de l’humus du sol. Trav. Inst. des Sols Dokutchaev, 38, 32 p Yamada E, Doi K, Okano K and Fuse Y (2000) Simultaneous determinations of the concentration and molecular weight of humic substances in environmental water by gel chromatography with a fluorescence detector. Analytical Sci., 16, 125–132 Yu-Ping Chin, Alken G, and O’Loughlin E (1994) Molecular weight, polydispersity, and spectroscopic properties of aquatic humic substances. Environ. Sci. Technol., 28, 1853–1858 Wershaw RL and Aiken GR (1985) Molecular size and weight measurements of humic substances. In : Humic substances in soil, sediment and water., Aiken GR, McKnight DM, Wershaw RL and MacCarthy P ed. Wiley New York, 477–492 Wershaw RL (1989) Sizes and shapes of humic substances by scattering techniques. In : Humic substances II., Hayes MHB, MacCarthy P, Malcolm RL and Swift RS ed. Wiley New York, 545–559

Functional Group of Humic Compounds
De Nobili M, Contin M and Leita L (1990) Alternative method for carboxyl group determination in humic substances. Can. J. Soil Sci., 70, 531–536 Fritz JS, Yamamura SS and Bradford EC (1959) Determination of carbonyl compounds. Anal. Chem., 31, 260–263 Glebko LI, Ulkina JU and Maximov OB (1970) A semi-micro method for the determination of quinoid groups in humic acids. Microchim. Acta., 1247–1254 Rosell RA, Allan AL, Agullo E and Gelos B (1972) Estudio potentiometrico del humus. II. Determinacion de varios tipos de acidez (o grupo funcionales) de acidos humicos de suelos de la provincia de Buenos Aires, Argentina. Turrialba, 22, 327–332

446

Organic Analysis

Schnitzer M and Gupta UC (1965) Determination of acidity in soil organic matter. Soil Sci. Soc. Am. Proc., 29, 274–277 Schnitzer M and Skinner SIM (1965) Organo-metallic interactions in soils : 4.Carboxyl and hydroxyl groups in organic matter and metal retention. Soil Sci., 99, 278–284 Wright JR and Schnitzer M (1959) Oxygen-containing functional groups in the organic matter of a Podzol soil. Nature, London, 184, 1462–1463

Spectrometric Characterizations UV–Visible, IR, Fluorescence, ESR Spectrometries
Lapen AJ and Seitz WR (1982) Fluorescence polarization studies of the conformation of soil fulvic acid. Anal. Chim. Acta., 134, 31–38 Bachelier G (1981) Etude spectrographique de la fluorescence des acides humiques et des acides fulviques de divers sols. Cah. ORSTOM sér. Pédol., 18, 129–145 Chen Y, Senesi N and Schnitzer M (1977) Information provided on humic substances by E4:E6 ratios. Soil Sci. Soc. Am. J., 41, 352–358 Ghosh K and Schnitzer M (1979) UV and visible absorption spectroscopic investigations in relation to macromolecular characteristics of humic substances. J. Soil Sci., 30, 735–745 Ghosh K and Schnitzer M (1980a) Fluorescence excitation spectra of humic substances. Can J. Soil Sci., 60, 373–379 Goldberg MC, Cunningham KM and Weiner ER (1987) The use of isosbestic points in the fluorescence excitation spectrum of humic acid to calculate the dissociation constant. Can. J. Soil Sci., 67, 715–717 Kononova MM (1966) Soil organic matter, its nature, its role in soil formation and in soil fertility, 2nd English ed. Pergamon Oxford, 544 p Lévesque M (1972) Fluorescence and gel filtration of humic compounds. Soil Sci., 113, 346–353 MacCarthy Pet Rice JA (1985) Spectroscopic methods (other than NMR) for determining the functionality in humic substances. In: Humic substances in soil, sediment and water., Aiken GR, McKnight DM, Wershaw RL and MacCarthy P ed. Wiley New York, 527–559 Nayak DC, Barman AK, Varadachari C, Ghosh K (1985) Fluorescence excitation spectra of humic acids. J. Indian Soc. Soil Sci., 33, 785–787 Rex RW (1960) Electron paramagnetic resonance studies on stable free radicals in lignins and humic acids. Nature, 188, 1185–1186 Riffaldi R and Schnitzer M (1972) Effects of divers experimental conditions on ESR spectra of humic substances. Geoderma, 8, 1–10 Russel JD and Anderson HA (1977) Comment on “spectroscopie infra-rouge de quelques fractions d’acides humiques obtenues sur Sephadex” Plant Soil, 48, 547–548

Characterization of Humic Compounds

447

Schnitzer M and Skinner SIM (1969) Free radicals in soil humic compounds. Soil Sci., 108, 383–390 Saab SC and Martin–Neto L (2004) Studies of Semiquinone Free Radicals by ESR in the Whole Soil, HA, FA and humin substances. J. Braz. Chem. Soc., 15, 34–37 Senesi N (1990) Application of ESR spectroscopy in soil chemistry. Adv. in Soil Sci., 14, 77–129

Nuclear Magnetic Resonance
Barron PF, Wilson MA, Stephens JF, Cornell BA and Tate KR (1980) Crosspolarization 13C NMR spectroscopy of whole soils. Nature, 286, 585–587 Brown JK and Ladner WR (1960) A study of the hydrogen distribution in coallike materials by high resolution NMR spectroscopy (II). Fuel, 39, 87–96 Conte P, Piccolo A, van Lagen B, Buurman P and de Jager PA (1997a) Quantitative differences by liquid- and solid-state 13C NMR spectroscopy. Geoderma, 80, 339–352 Conte P, Piccolo A, van Lagen B, Buurman P and de Jager PA (1997b) Quantitative aspects of solid-state 13C NMR spectra of humic substances from soils of volcanic systems. Geoderma, 80, 327–338 Conte P, Spaccini R and Piccolo A (2004) State of the art of CPMAS C-13NMR spectroscopy applied to natural organic matter. Progress in Nuclear Magnetic Resonance Spectroscopy, 44, 215–223 Doddrell D, Glushko V and Allerhand A (1972) Theory of the Nuclear Overhauser Enhancement and 13C 1H dipolar relaxation in protondecoupled carbon-13 NMR spectra of macromolecules. J. Chem. Physics, 56, 3683–3689 Freeman R, Hill HDW, Kaptein R (1972) Proton-decoupled NMR spectra of carbon-13 with the nuclear Overhauser effect suppressed. J. Magn. Res., 7, 327–329 Fründ R. and Lüdemann HD (1989) The quantitative analysis of solution and CPMAS-C13 NMR Spectra of humic material. Sci. Total Environ., 81/82, 157–168 Gerasimowicz WV and Byler DM (1985) Carbon-13 CPMAS NMR and FTIR spectroscopic studies of humic acids. Soil Sci., 139, 270–278 Kinchesh P, Powlson DS and Randall EW (1995) 13C NMR studies of organic matter in whole : a case study of some Rothamsted soils. Eur. J. Soil Sci., 46, 139–146 Lentz H, Ludemann HD and Ziechmann W (1977) Proton resonance spectra of humic acids from the solum of a podzol. Geoderma, 18, 325–328 Lowe IJ (1959) Free induction decay of rotating solids. Phys. Rev. Lett., 2, 285–287

448

Organic Analysis

Moser A, Lefebvre B (2004) Identifying Residues in Natural Organic Matter through Spectral Prediction and Pattern Matching of 2-D NMR datasets. Magn. Resonance Chem., 42, 14–22 Newman RH and Tate KR (1984) Use of alkaline soil extracts for 13C NMR characterization of humic substances. J. Soil Sci., 35, 47–54 Newman RH, Tate KR, Barron PF and Wilson MA (1980) Towards a direct, non-destructive method of characterising soil humic substances using 13C-NMR. J. Soil Sci., 31, 623–631 Pines A, Gibby MG and Waugh JS (1973) Proton enhanced NMR of dilute spins in solids. J. Chem. Phys., 59, 569–590 Preston CM and Schnitzer M (1984) Effects of chemical modifications and extractants on the 13C NMR spectra of humic materials. Soil Sci. Soc. Am. J., 48, 305–311 Schaefer J and Stejskal EO (1976) C-13 NMR of polymers spinning at the magic angle, J. Am. Chem. Soc., 98, 1031–1032 Schnitzer M and Barton DHR (1963) A new approach to the humic acid problem. Nature, London 198, 217–219 Schnitzer M and Neyroud JA (1974) The chemistry of high molecular weight fulvic acid fractions. Can. J. Chem., 52, 4123–4132 Simpson AJ, Kingery WL, Williams A, Golotvin S, Kvasha M, Kelleher BK, Simpson AJ, Tseng L, Spraul M, Brauman U, Kingery WL, Kelleher B, Simpson MJ (2004) The application of LC-NMR and LC-SPE-NMR for the separation of Natural Organic Matter. The Analyst, 129:1216– 1222 Smernik RJ and Oades MJ (2000) The use of spin counting for determining quantitation in solid state 13C NMR spectra of natural organic matter. 1. Model systems and the effects of paramagnetic impurities. Geoderma, 96, 101–129 Smernik RJ and Oades MJ (2000) The use of spin counting for determining quantitation in solid state 13C NMR spectra of natural organic matter. 1. HF-treated soil fractions. Geoderma, 96, 159–171 Smernik RJ and Oades MJ (2003) Spin accounting and RESTORE, two new methods to improve quantitation in solid-state 13C NMR analysis of soil organic matter, Eur. J. Soil. Sci., 54, doi:10.1046/j.1365–2389. 2003.00497.x Steelink C, Wershaw RL, Thorn KA and Wilson MA (1989) Application of liquid-state NMR spectroscopy to humic substances. In : Humic substances II, Hayes MHB, MacCarthy P, Malcolm RL and Swift RS ed. Wiley New York, 281–338 Tate RL (1998) Humic substances and organic matter in soil and water environments: characterization, transformations and interactions. Soil Sci., 163, 675–676 Thorn KA and Mikita MA (2000) Nitrite fixation by humic substances: nitrogen15 nuclear magnetic resonance. Evidence for potential intermediates in chemodenitrification. Soil Sci. Soc. Am. J., 64, 568–582

Characterization of Humic Compounds

449

Wilson MA (1981) Applications of nuclear magnetic resonance spectroscopy to the study of the structure of soil organic matter J. Soil Sci., 32, 167–186 Wilson MA (1987) Techniques and applications of NMR spectroscopy in geochemistry and soil science., Pergamon, Oxford Wilson MA (1989) Solid-state NMR spectroscopy of humic substances, basic concepts and techniques. In : Humic substances II, Hayes MHB, MacCarthy P, Malcolm RL and Swift RS ed. Wiley and Sons, 309–338 Wilson MA, Jones AJ and Williamson B (1978) NMR spectroscopy of humic materials. Nature, London, 276, 487–489

Methods of Characterization by Fragmentation
Anderson HA, Hepburn A and Sim A (1978) Ether-soluble hydrolysis products in humic and fulvic acids. J Soil Sci., 29, 84–87 Bracewell JM, Haider K, Larter SR and Schulten HR (1989) Thermal degradation relevant to structural studies of humic substances. In : Humic substances II, Hayes MHB., MacCarthy P, Malcolm RL and Swift RS ed., Wiley New York, 181–222 Chakrabartty SK, Kretschmer HO and Cherwonka S (1974) Hypohalite oxidation of humic acids. Soil Sci., 117, 6 Chen Y, Senesi N and Schnitzer M (1978a) Chemical degradation of humic and fulvic acids extracted from mediterranean soils. J. Soil Sci., 29, 350– 359 Chen Y, Senesi N and Schnitzer M (1978b) Chemical and physical characteristics of humic and fulvic acids extracted from soils of the mediterranean region. Geoderma, 20, 87–104 Cheshire MV, Cranwell PA and Haworth RD (1968) Humic acid – III. Tetrahedron, 24, 5155–5167, Pergamon UK Griffith SM and Schnitzer M (1989) Oxidative degradation of soil humic substances. In : Humic substances II, Hayes MHB, MacCarthy P, Malcolm RL and Swift RS ed., Wiley New York, 69–98 Griffith SM and Schnitzer M (1975) Oxidative degradation of humic and fulvic acids extracted from tropical volcanic soils. Can. J. Soil Sci., 55, 251– 267 Griffith SM and Schnitzer M (1976) The alkaline cupric oxide oxidation of humic and fulvic acids extracted from tropical volcanic soils, Soil Sci., 122, 191–201 Hansen EH and Schnitzer M (1967) Nitric acid oxidation of Danish illuvial organic matter. Soil Sci. Soc. Am. Proc., 31, 79–85 Hansen EH and Schnitzer M (1969) Zn–dust distillation and fusion of a soil Humic and fulvic acid. Soil Sci. Soc. Am. Proceed., 33, 29 Hatcher PG, Dria KJ, Kim S., Frazier, SW (2001) Modern Analytical Studies of Humic substances. Soil Sci., 166, 770–794 Hayes MHB and O’Callaghan MR (1989. Degradations with sodium sulfide and with phenol. In : Humic substances II, Hayes MHB, MacCarthy P, Malcolm RL and Swift RS ed., Wiley New York, 143–180

450

Organic Analysis

Jackson MP, Swift RS, Posner AM, Knox JR (1972) Phenolic degradation of humic acid. Soil Sci., 114, 75–78 Kahn SU and Schnitzer M (1971a) The permanganate oxidation of methylated and unmethylated humic acids extracted from Solonetz, Solod and Chernozem Ah horizons. Israel J. Chem., 9, 667–677 Khan SU and Schnitzer M (1971b) Further investigations on the chemistry of fulvic acid, a soil humic fraction. Can. J. Chem., 49, 2302–2309 Kahn SU and Schnitzer M (1972a) Permanganate oxidation of humic acids, fulvic acids, and humins extracted from Ah horizons of a black chernozem, a black solod and a black solonetz soil. Can. J. Soil Sci., 52, 43–51 Kahn SU and Schnitzer M (1972b) Permanganate oxidation of humic acids extracted from a GrayGrey wooded soil under different cropping systems and fertilizer treatments. Geoderma, 7, 113–120 Kimber RWL and Searle PL (1970) Pyrolysis gas chromatography of soil organic matter.1.Introduction and methodology, Geoderma, 4, 47–55 Kodama H and Schnitzer M (1970) Kinetics and mechanism of the thermal decomposition of fulvic acid. Soil Sci., 109, 265–271 Martin F (1976) Effects of extractants on analytical characteristics and pyrolysis gas chromatography of podzol fulvic acids. Geoderma, 15, 253–265 Mendez J and Stevenson FJ (1966) Reductive cleavage of humic acids with sodium amalgam. Soil Sci., 102, 85 Meuzelaar HLC, Haider K, Nagar BR and Martin JP (1977) Comparative studies of pyrolysis mass spectra from melanins of soil fungi, model phenolic polymers and humic acids from soil, peat and composted straw. Geoderma, 17, 239–252 Meuzelaar HLC, Posthumus MA, Kistemaker PG and Kistemaker J (1973) Curie point pyrolysis in direct combination with low voltage electron impact ionization spectrometry. Anal.Chem., 45, 1546–1549 Morrison RI (1963) Products of the alkaline nitrobenzene oxidation of soil organic matter. J. Soil Sci., 14, 2 Neyroud JA and Schnitzer M (1974) The exhaustive alkaline cupric oxide oxidation of humic and fulvic acid, Soil Sci. Soc. Am. Proc., 38, 907– 913 Neyroud JA and Schnitzer M (1975) The alkaline hydrolysis of humic substances. Geoderma, 13, 171–188 Ogner G (1975) Oxidation of nonhydrolyzable humic residue and its relation to lignin, Soil Sci., 116, 93–100 Ouchi K and Brooks JD (1967) The isolation of certain compounds from depolymerized brown coal. Fuel, 46, 367–377 Parsons JW (1989) Hydrolytic degradations of humic substances. In : Humic substances II, Hayes MHB, MacCarthy P., Malcolm R.L. and Swift RS ed., Wiley New York, 99–120 Piper TJ and Posner AM (1972) Sodium amalgam reduction of humic acid (I and II). Soil Biol. Biochem., 4, 513–531

Characterization of Humic Compounds

451

Saiz-Jimenez C, Haider K and Meuzelaar HLC (1979) Comparisons of soil organic matter and its fractions by pyrolysis mass-spectrometry. Geoderma, 22, 25–37 Schnitzer M and Hoffmann I (1964) Pyrolysis of soil organic matter. Soil Sci.Soc.Proced., 520–525 Schnitzer M and Skinner SIM (1974) The peracetic acid oxidation of humic substances, Soil Sci., 118, 322–331 Schulten HR (1996) Direct pyrolysis–mass spectrometry of soils: a novel tool in agriculture, ecology, forestry and soil science. In Mass spectrometry of soils, Boutton TW and Yamasaki S-i ed., Marcel Dekker New York Stevenson FJ and Mendez J (1967) Reductive cleavage products of soil humic acids. Soil Sci., 103, 383 Stevenson FJ (1989) Reductive cleavage of humic substances. In : Humic substances II, in search of structure, Hayes MHB, MacCarthy P, Malcolm RL and Swift RS ed., Wiley New York, 122–142

Other methods (microscopy, X-ray, electrochemistry, etc.)
Bachelier G (1983) Figures de dessication des acides humiques, Rapport multig. ORSTOM, 14 p Chen Y and Schnitzer M (1976) Scanning electron microscopy of a humic acid and of a fulvic acid and its metal and clay complexes. Soil Sci. Soc. Am. J., 40, 682–686 Chen Y, Senesi N and Schnitzer M (1976) Chemical and physical characteristics of humic and fulvic acids extracted from soils of the mediterranean region. Geoderma, 20, 87–104 Shinozuka N and Hayano S (1987) Polarographic characterization of humic acid. Soil Sci., 143, 157–161 Stevenson FJ and Schnitzer M (1982) Transmission electron microscopy of extracted humic and fulvic acids, Soil Sci., 133, 334–345 Tan KH (1985) Scanning electron microscopy of humic matter as influenced by methods of preparation. Soil Sci. Soc. Am. J., 49, 1185–1191 Defosse C and Rouxhet PG (1980) Introduction to X-ray photoelectron spectroscopy. In: Advanced chemical methods of soil and clay mineral research., J.W Stucky and W.L. Banwart ed., Reidel, Dordrecht, 165– 204 Goodman BA and Cheshire MV (1979) A Mössbauer spectroscopic study of the effect of pH on the reaction between iron and humic acid in aqueous media. J. Soil Sci., 30, 85–91

13

Measurement of Non-Humic Molecules

13.1 Introduction

13.1.1 Non-Humic Molecules Soil organic matter probably contains the majority of biochemical compounds synthesized by living organisms (Stevenson 1982). In addition to humic molecules, whose quantification (described in Chap. 11) is technically simpler than the structural characterization of molecules (cf. Chap. 12), a large number of other molecules of known structure are also present in soils. The most abundant can be classified in three main groups: 1. Nitrogenous molecules, which are often studied by their fragmentation products during acid hydrolysis. 2. Polysaccharides, mostly not nitrogenous molecules, which are also often studied by their fragmentation products. 3. The lipid fraction, which are sometimes called soil bitumens and contain a range of molecules extractable with the solvents used for fats; this fraction also contains many of the organic pollutants and xenobiotic residues that can contaminate the soil. Titration techniques for (1) nitrogenous molecules are described in Chap. 14. This chapter describes the main titration techniques for the second and third types of molecules. 13.1.2 Soil Carbohydrates Total Composition Sugars (also improperly called carbohydrates because of an empirical formula corresponding to Cn(H2O)n) account for 5–25% of soil organic

454

Mineralogical Analysis

matter. Aside from traces of free sugars which can be extracted from the soil by water, soil carbohydrates are components of polysaccharides. It is practically impossible to isolate soluble fractions of polysaccharides for identification (Cheshire 1979). Consequently it is difficult to know if sugars come from a heterogeneous mixture of polysaccharides or from a single particularly complex polysaccharide. Eight neutral carbohydrates can be identified by hydrolysis of soils. These can be classified in three main groups: hexose, deoxyhexose and pentose sugars. Two acid sugars (uronic acids) and two basic sugars (hexosamines) can also be identified. Ranked in descending order, hexose sugars account for from 12–4% of organic matter and include glucose, galactose and mannose. Uronic acids account for 1–5% of soil organic matter and contain roughly equal parts of galacturonic and glucoronic acids. Pentose sugars are present in smaller quantities and contain mainly arabinose and xylose as well as traces of ribose. Fucose and rhamnose deoxyhexoses are found in similar concentrations to pentose sugars. Amino sugars (cf. Chap. 14) are present still in lower concentrations in the soil; their chief components are hexosamines in the form of galactosamine and glucosamine. Traces of other sugars can also be found in the soil: four methyl sugars, two alcohol sugars (inositol and manitol), two hexose sugars (fructose and sorbose), a pentose sugar (deoxyribose) and a hexosamine sugar (N-acetyl glucosamine). Sugars and Types of Soils Many authors have tried to characterize soils by quantification of the sugars that result from acid hydrolysis of polysaccharides. Folsom et al. (1974) found an almost linear relationship between total carbohydrate content and soil organic carbon content; however a curvature appeared for the range of very organic horizons which contained a lower proportion of sugars. These authors reported that grassland soils contained more pentoses and less hexoses than forest soils and also that the proportion of mannose increased with soil depth, indicating greater stability of this sugar. Singhal and Sharma (1985) reported that total carbohydrates and organic carbon contents of forest soils varied in the same way whatever the tree cover. They observed no difference in the relative proportions of these sugars. MacGrath (1973) also found a very constant value for the relative composition of sugars in 38 Irish grassland soils.

Non-humic Molecules

455

Cheshire and Anderson (1975) observed a higher quantity of total sugar in cultivated soils than in uncultivated soils but the relative proportions were the same. Distribution and Origin Another aspect of the problem is linked to the distribution of polysaccharides in the fractions of soil organic matter. Diluted alkaline solutions are the best reagent for polysaccharide extraction although the majority of the polysaccharides remain associated with humin residues. Acidification of alkaline extracts precipitates humic acids and the majority of the polysaccharides remain in the acid soluble fraction of fulvic acid (Cheshire 1979; Barriuso et al. 1985). Bagautdinov et al. (1984) separated one fraction from a fulvic acid solution which contained mainly polysaccharides with a molecular weight of 27,000–28,000. Many authors have tried to determine whether soil sugars are of microbial or plant origin. This is not a simple task since none of the main sugars can be classified exclusively as being of plant or microbial origin (Cheshire 1979). There are more similarities than differences in sugars between humic acids and fungi melanins (Coelho et al. 1988). Incubation experiments using labelled glucose led to labelling of all sugars and, to a lesser extent, of amino acids. Deoxyhexoses have been identified as the most stable synthesized sugars (Cheshire 1979). With an increase in incubation time, an increasingly large proportion of labelled carbon was found in humin (Guckert et al. 1971). Hexose sugars are the main sugars synthesized by soil micro-organisms (Oades 1974). François (1988) and Murayama (1983, 1984, 1988) reported that they had been unable to identify a specific origin for glucose, galactose and ribose but that xylose and arabinose are primarily of plant origin, while rhamnose, mannose and fucose are very often synthesized by microorganisms but are also present in root exudates of various plants. François (1988) showed that total sugars are primarily concentrated in fresh roots and in the 5–25 µm soil particle fraction. A marked relative reduction in xylose content and an increase in mannose and rhamnose contents were measured in the finer fractions. Feller (1991) used the xylose:mannose ratio as an indicator of the microbial or plant origin of the organic matter. This ratio had the lowest value (0.5–2) in the organic matter of the clay-organic compartment <2 µm; its value ranged between 1 and 3 in the 2–20 µm fraction and between 5 and 10 in the >20 µm fraction (cf. Chap. 9).

456

Mineralogical Analysis

Principle of Titration Free sugars can be titrated on aqueous extracts of soils. Polysaccharide titration has three main stages: – Acid hydrolysis of polysaccharides; – Possible purification of the hydrolysate; and – Titration of the hydrolysate. Titration of free or hydrolysed sugars uses two types of techniques: – Global colorimetric methods for reducing sugars and – Chromatographic methods allowing the measurement of each individual sugar. 13.1.3 Soil Lipids Soil lipids are rather complex mixtures of compounds one of whose common characteristics is solubility in a range of organic solvents or mixtures of solvents. This fraction includes groups such as free fatty acids, hydrocarbons, polar or non-polar lipids, steroids, waxes and resins. The majority of lipids can be classified in three main groups: fats, waxes and resins. The resins are the most polar compounds and are thus most soluble in methanol and ethanol, and this property can be used to separate them. Lipids are present in the soil in smaller quantities than nitrogenous compounds and polysaccharides. According to Stevenson (1982) they account for 2–6% of organic matter) and according to Jambu et al. (1978) up to 20% in certain soils. Lipids are generally present in higher concentrations in acid soils. The most fertile soils are generally poor in lipids. The presence of lipids may even be related to the old concept of soil sickness (Stevenson 1966) and depend on humus content, soil aeration and texture (Jambu et al. 1978). The extraction of soil lipids may be complicated by the fact they bond to a varying extent with other organic or inorganic compounds in the soil. Most pesticides and other organic pollutants of the soils are also extracted with the lipid fraction.

Non-humic Molecules

457

13.1.4 Pesticides and Pollutants Most of the products used in food and agricultural chemistry can be found in soils. Organic pollutants that result from the breakdown of products in the environment can be extremely varied, although the majority belong to the following main groups: – Polychlorinated biphenyls (PCBs) are families of chlorinated pollutants including about 30 compounds; traces of PCB can be detected in a very large number of substrates; – Polynuclear aromatic hydrocarbons (PAH) include 15–20 compounds (e.g. naphthalene, phenantrene, anthracene or pyrene) that are resistant to degradation; and – Dioxins are also residues of degradation but are present in more limited quantities. Pesticides are generally not very stable molecules used in agriculture in a wide range of products like insecticides, acaricides, nematocides, repellents, fungicides, herbicides and poisons. They comprise a very large number of compounds. Although there is a general trend towards molecules that are less and less toxic for humans, and more and more degradable, there is also an increase in the total amount used as a result of gene mutations and of the development of resistance in living organisms. Calvet et al. (2005) summarized current knowledge about pesticides and their agronomic and environmental consequences for soils. More than one thousand compounds are sold as pesticides and these are difficult to classify. Some of the main families are: – Organochlorinated products including the first historically sold synthetic molecules such as DDT, dieldrin or lindane; these products are less used today because of their toxicity and low rate of degradability; however, they are still found in the environment since they accumulated in the lipids of living organisms; nowadays new halogenated molecules are used; – Organophosphorous products were often used to replace the first organochlorinated products; – Triazine herbicides; – Acid herbicides like chlorophenoxyacetic acids, picloram or dicamba; – Carbamates and thiocarbamates, which are often used as systemic insecticides; and – Pyrethrinoids, which are synthetic molecules derived from natural pyrethrins and are commonly used as insecticides because they are not exchanged in the blood of human beings or warm-blooded animals.

458

Mineralogical Analysis

13.2 Classical Techniques
13.2.1 Acid Hydrolysis of Polysaccharides Principle of the Technique and Main Difficulties Hot diluted acids are capable of completely hydrolyzing polysaccharides but in most soils they only hydrolyze about 75%. To achieve complete hydrolysis of polysaccharides, e.g. cellulose or soil polysaccharides, a preliminary treatment with a more concentrated acid is required (Cheshire 1979). During hydrolysis with diluted acids, release of hexose sugars increases with an increase in the concentration of the acid. Alone, diluted acids are generally not appropriate for quantitative titration of hexose sugars. Pentose sugars are more easily released than hexose sugars but they are also more easily destroyed during acid hydrolysis and, depending on the soil, they may be more efficiently titrated by hydrolysis in more diluted medium. Ivarson and Sowden (1962) were the first to recommend cold pretreatment with 12 mol L–1 sulphuric acid (72%) followed by dilution of the acid to 0.5 mol L–1 and heating at reflux. In samples of litter, this attack released almost three times more hexoses than hydrolysis without pretreatment, but the release of pentoses was reduced by approximately 20% in the case of conifer litter. Gupta and Sowden (1965) confirmed on four soils a very positive effect of pretreatments for the titration of hexoses and pentoses, except in some cases where deoxyhexoses were partly destroyed. Cheshire and Mundie (1966) optimized the time of cold pretreatment with 12 mol L–1 sulphuric acid. The sugars measured by orcinol colorimetry reached maximum towards 16 h then decreased, whereas those measured by anthrone colorimetry (hexoses) continued to increase up to 40 h. The length of pretreatment thus influenced the time of attack at reflux with the 0.5 mol L–1 sulphuric acid solution necessary for the maximum release of sugar. After 16 h cold pretreatment, 5 h of hot attack were sufficient to reach maximum, whereas after 2 h pretreatment, nearly 20 h of attack were necessary. Finally, these authors concluded that there is no perfect method of hydrolysis enabling the complete release of glucose without partially destroying pentoses or deoxyhexoses. The 16 h treatment with 12 mol (H2SO4) L–1 at 20°C followed by heating for 5 h at reflux with 0.5 mol (H2SO4) L–1 was recommended because in several cases it resulted in the highest rate of release of glucose. This reason is not very convincing in the case of studies on the origin of sugars in the soil where glucose is not representative of a sugar of plant or microbial origin (cf. Sect. 13.1.1).

Non-humic Molecules

459

Oades et al. (1970) focussed on hydrolysis conditions for chromatographic titration of sugars. For a soil with no plant fragments, they found the best extracted sugar contents by direct attack at reflux for 1 h with a 5 mol (½H2SO4) L–1 solution. However, this attack released less glucose than another technique similar to that of Cheshire and Mundie (16 h in 13 mol (H2SO4) L–1 cold reagent followed by heating at reflux for 2 h in 0.5 mol (H2SO4) L–1). But Oades’ technique released a higher quantity of other sugars, particularly xylose, rhamnose and fucose. However in a sandy silt soil, there is a risk in significant breakdown of sugars when the duration of hydrolysis exceeds 20 min for xylose, 40 min for arabinose and about 1 h for other sugars (not the same risk in the method with pretreatment). In fact for total sugar contents, the results of the method with heating at reflux for 20 min in 5 mol (½H2SO4) L–1 were not very different from those obtained using cold pretreatment, except for samples rich in materials where method with pretreatment is recommended. The Oades’ method provided more pentoses and deoxyhexoses but much less glucose. To release glucose, a longer attack with 5 mol (½H2SO4) L−1 was necessary but with the risk of destroying pentose and deoxyhexose sugars. To obtain maximum rates for glucose and other sugars simultaneously Oades et al. recommended heating at reflux for 20 min with 5 mol (½H2SO4) L–1 reagent then filtration followed by 16 h of cold maceration in 26 mol (½H2SO4) L–1 reagent and heating at reflux for 5 h in 1 mol (½H2SO4) L–1 reagent. Initial heating at reflux with 5 mol (½H2SO4) L–1 reagent for only 20 min may be sufficient for not very organic soils but there is a risk of underestimating glucose. The majority of more recent studies did not continue to optimize hydrolysis conditions; Coelho et al. (1988), Murayama (1987), Cheshire and Griffiths (1989) used the method of Oades et al. (1970) with hydrolysis in two stages. They sometimes distinguished sugars released by heating at reflux for 20 min with 5 M H2SO4, called non-cellulose sugars, from sugars released by later hydrolysis (heating at reflux for 5 h with 1 N H 2SO4 preceded by maceration in 26 N H2SO4), called cellulose sugars. Arschad and Schnitzer (1987) and Baldock et al. (1987) used the method developed by Spiteller (1980) which is similar to that of Cheshire and Mundie (1966) with respect to the conditions of hydrolysis: maceration for 16 h with 26 N H2 SO4 followed by boiling at reflux for 5 h with N H 2 SO4. Singhal and Sharma (1985) used the method of Gupta (1967): 2 h of extraction at low temperature with 72% H2SO4 then

460

Mineralogical Analysis

heating at reflux for 16 h after dilution. Benzing-Purdie and Nikiforuk (1989) used a simpler technique for hydrolysis at 105°C for 18 h in 2 N H 2 SO4 and compared it with the method of Cheshire and Mundie (1966). Guckert (1973) also used a similar technique with hydrolysis with 3 mol (½H 2 SO4) L–1 acid at 80°C for 24 h after extraction. Without cold pretreatment, with 4.5 h hydrolysis, titration of sugars by both anthrone and ferricyanide colorimetry reached maximum with a 5 N acid concentration and a temperature of 100°C (in closed flasks) and displayed a much less significant influence of the acid concentration than of temperature (Pansu 1992). Cold pretreatment increased the quantities of sugars released and the optimum temperature for the subsequent attack moved to between 70 and 90°C with a weaker acid (1 mol (½H2SO4) L–1, in agreement with the results of Cheshire and Mundie (1966). Pansu (1992) also showed good resistance to the degradation of known quantities of monosaccharides added in the attack solutions. Commercial crystallized cellulose showed good resistance to concentrated acids at 105°C whatever the concentration of the sulphuric acid, and even to hydrochloric acid. Cellulose hydrolysis only became significant (88%) with the method including a cold pretreatment with the 26 mol (½H2SO4) L–1 reagent. Hydrolysis of commercial cellulose released not only glucose (66– 86% of the total), but also galactose (10–30%), mannose (1–2%), xylose (approximately 1%) and ribose (approximately 1%). Equipment and Reagents – Inorganic purified water: demineralized water purified on activated carbon (standard Millipore) or water distilled after reflux attack in the presence of a little potassium permanganate and sulphuric acid. – Six round bottom boiling flasks (100 mL) with ground stopper or six flasks with PTFE joint screw stopper (100 mL). – Set of six condensors (only for the method with reflux heating). – Büchner funnels with standard GF/A glass fibre filters φ 4 cm (Whatman No. 1820 042, or similar). – 13 mol L–1 concentrated sulphuric acid solution: add 278 mL inorganic water in a 1 L volumetric flask; carefully add while agitating and cooling in a cold water bath, 722 mL of concentrated reference grade sulphuric acid; complete to 1 L. – 2.5 mol (H2SO4) L–1 (5 N) sulphuric acid solution: proceed as given earlier but with 139 mL of acid and complete to 1 L with inorganic water.

Non-humic Molecules

461

Procedure Section 13.2.1 earlier mentions that our current state of knowledge makes it difficult to recommend a precise standardized attack that is valid for all soils. The conditions of hydrolysis need to be adjusted to the type of study required. Based on the review of the literature in Sect. 13.2.1, the three alternative procedures listed later can all be used. Attack Favouring the Measurement of Pentoses Put 5 g of soil and 40 mL of 2.5 mol L–1 sulphuric acid in a 100 mL round bottom boiling flask. Connect to the condensor and heat at reflux for 20 min (alternatively heat in a flask with a screw cap at 100°C for 20 min). Leave to cool and filter on a Büchner funnel on glass fiber filter or centrifuge. Total Attack Add the residue of the previous attack (with glass fibre filter) and 2 mL of 13 mol (H2SO4) L–1 solution in the same 100 mL round bottom flasks used earlier. Stop the flasks and leave at ambient temperature for 16 h (overnight). Carefully add 50 mL inorganic water while cooling. Connect the flasks to the condensors and heat at reflux for 5 h (alternatively, heat at 90°C in a closed bottle). Filter as mentioned earlier on glass wool. For the analysis of total sugars, the filtrates described in “Attack Favouring the Measurement of Pentoses” earlier and in “Total Attack” can be mixed. The analysis of each hydrolysate releases (a) sugars that are easily hydrolysable (mainly pentoses) in the hydrolysate described in “Attack Favouring the Measurement of Pentoses”, (b) sugars like cellulose that are difficult to hydrolyse in the hydrolysates described in “Total Attack”. Simplified Total Attack In a 100 mL round bottom flask place 5 g of soil sample and 2 mL of 13 mol (H2SO4) L–1 solution. Stop the flask and leave at the ambient temperature for 16 h (overnight). Carefully add 50 mL inorganic water while cooling. Connect the boiling flasks to the condensors and heat at reflux for 5 h (alternatively, heat at 90°C in a closed bottle). Filter as mentioned earlier on glass wool or centrifuge.

462

Mineralogical Analysis

13.2.2 Purification of Acid Hydrolysates Principle of the Technique and Main Difficulties Particularly for chromatographic titration, the hydrolysates must be purified to remove a significant amount of inorganic compounds especially sulphate ions but also inorganic dissolved solids (particularly iron and aluminium) by the acid attack of the soil. Neutralization of the extracts involves the precipitation of aluminium and iron hydroxides. If neutralization is carried out with a carbonate or hydroxide belonging to the alkaline earths (Ca, Sr or Ba) sulphates can also be precipitated. The risk of error lies in the possible adsorption of neutral sugars by the precipitates. With barium hydroxide or carbonate, neutral sugars and barium salts of uronic acids can be recovered by filtration and by washing the precipitates with water, sometimes by washing with hot water or by ethanol extraction with a Sohxlet apparatus (Cheshire and Mundie 1966). Sodium bicarbonate has also been used for neutralization of the acid extracts with methanol extraction of the sodium sulphate after desiccation (Oades 1967). Calcium carbonate enables neutralization by eliminating sulphates together with most coloured organic matter (Brink et al. 1960). Oades et al. (1970) preferred strontium carbonate to bring the hydrolysates towards pH 7. When the solutions were left to stand, these authors observed coprecipitation of brown organic matter and iron complexes. After filtration and dry evaporation of the filtrate in a rotary evaporator, sugars were dissolved in 4–5 fractions of 2–4 mL methanol with filtration. The extracts can be also purified by successive passage on cation and anion resin which also separates neutral sugars and uronic acids. Alternatively, columns filled with a mixture of coal and celite enable sugars to be eluated with 50% ethanol in water after the salts are rinsed in water (Cheshire 1979). Tests were carried out at IRD1 (unpublished data) to compare seven methods of purification: – Water elution on MB1 amberlite resin, – Methanol elution on MB1 amberlite resin, – Neutralization with strontium carbonate, – Neutralization with sodium bicarbonate, – Neutralization with barium carbonate,

1

IRD, Institute of Research for Development (ex-Orstom), BP 64501, 911 Avenue of Agropolis, 34394 Montpellier Cedex 5, France.

Non-humic Molecules

463

– Neutralization with barite and – Neutralization with soda. The method which provided the best recovery of sugars was neutralization with strontium carbonate; this result is in agreement with that of Oades et al. (1970). Corrective Factors Studies using synthetic solutions enabled the effect of neutralization with strontium carbonate on the recovery of sugars to be identified (Pansu 1992). In three tests, the internal standard (myoinositol) was added before neutralization, and in three other tests, myoinositol was added before the borohydride reduction of the purified concentrated solutions. Each final solution was injected twice into the chromatograph. Calculations of the absolute contents showed: – The error due to the method of preparation of the samples was no higher than the error in the chromatographic measurement (F test). For most sugars, this error was about 1%. – A higher rate of adsorption of the internal standard on the precipitate than adsorption of the other sugars. Thus it is better to introduce the internal standard after purification of the extracts but before derivatization of sugars. – An average percentage of recovery of sugars of 85%, ranging from 71% for ribose to 91% for rhamnose. Ribose was always recovered in the smallest quantities. Based on these recovery percentages, we propose the following corrective factors: multiply the results of absolute contents of each sugar by 1.21 for rhamnose, 1.08 for fucose, 1.41 for ribose, 1.10 for arabinose and xylose, 1.25 for mannose, 1.21 for galactose and 1.15 for glucose. These factors take into account the response coefficients of each sugar compared to the internal standard. Equipment and Reagents – Six 100 mL beakers; – Magnetic stirrers; – Six 250 mL round bottom flasks for the rotary evaporator and – Strontium carbonate, M = 147.63 g. 2.2.4 Procedure Transfer the filtrates or centrifugation solutions described in Sect. 13.2.1 in 100 mL beakers, agitate on a magnetic stirrer. Add the quantity of

464

Mineralogical Analysis

strontium carbonate in powder form calculated according to the quantity of sulphuric acid to be neutralized plus 10%, i.e.: – Hydrolysis A: 1.1 × 40 × 2.5 × 147.63/1000 = 16.2 g – Hydrolysis C: 1.1 × 2 × 13 × 147.63/1000 = 4.2 g – Hydrolysis B: the same as hydrolysis c + 2 mL of 2.5 mol L–1 acid solution which impregnates the a residue, i.e. a total of 4.9 g. SrCO3 should be added to the solution slowly under agitation to avoid foam overflow. Leave in contact for at least 1 h under agitation. Check the pH is neutral or slightly basic (add a soda pellet if necessary). Check for precipitation of iron hydroxides which colour the precipitate and discolour the solution. Centrifuge and collect the supernatant in a round bottom boiling flask for later titration by gas chromatography (GPC) if required (cf. Sect. 13.2.4), and in a 100 mL volumetric flask if not. Rinse the residue with 25 mL methanol while stirring with a glass rod. Centrifuge again and add the supernatant to the previous solution. Repeat this operation once more. 13.2.3 Colorimetric Titration of Sugars Techniques Most colorimetric methods are based on one of the two following properties of sugars: their reducing power or, in strong acids, the formation of furfural-type compounds that react easily giving coloured derivatives. The methods can be classified in different categories corresponding to the measurement of total sugars, hexoses or pentoses. Colorimetry of Total Sugars Two titration techniques based on the reducing power of carbohydrates were tested (a) reduction of alkaline cupric salt solutions (Fehling’s liquor) resulting in Cu+ cooper complexes or (b) reduction of yellow ferricyanide solutions resulting in colourless ferrocyanides. The latter was considered preferable for the measurement of soil sugar and was automated by Cheshire and Mundie (1966). Three techniques use the other property based on furfural derivatives: total sugars with anthrone, phenol or orcinol. Anthrone produces a beautiful green–blue colour when it comes in contact with the furfural derivatives in the concentrated sulphuric acid. This reagent provides the best absorbance of complexes with deoxyhexose and hexose sugars (except for mannose). However the response to pentose sugars is

Non-humic Molecules

465

weaker and becomes undetectable when the anthrone content is above 0.05%. This method is also suspected of interfering with other organic matter as well as with iron and nitrates (Cheshire 1979). Phenol reacts with the furfural derivatives and results in a yellow colouration (Dubois et al. 1956). The similarity of this colour to that of soil hydrolysates could result in overestimation of total sugars with this reagent (McGrath 1973). Overestimated data were observed with direct phenol colorimetric analysis without purification of hydrolysates (Pansu 1992). However, Doutre et al. (1978) considered this method more satisfactory than the anthrone method. Orcinol or 3,5-dihydroxytoluene also reacts with the furfural derivatives, in this case with the advantage of providing enough similar responses for each sugar. It was used for soil hydrolysates by Bachelier (1966). Colorimetry of Hexose Sugars Although originally proposed for total sugars, anthrone is more commonly used to measure hexoses and deoxyhexoses of soil hydrohydrolysates. Ivarson and Sowden (1962) also proposed chromotropic acid for the measurement of hexoses. This reagent is not very susceptible to interference with pentose sugars and uronic acids and responded more strongly than anthrone on four samples of soil and litters. Colorimetry of Pentose Sugars Cheshire and Mundie (1966) used the orcinol-FeCl3 reagent described by Thomas and Lynch (1961) to measure the maximal release of pentose sugars during hydrolysis. In acetic acid, aniline also reacts with pentose sugars at ambient temperature resulting in a red colouration, with only slight interference by hexose sugars and uronic acids (Ivarson and Sowden 1962; Tracey 1950). Colorimetry of Deoxyhexose Sugars The yellow colour formed by heating the carbohydrate extract with cysteine in sulphuric acid medium is quoted as specific to deoxyhexose sugars; this reagent was used for soil hydrolysates by Cheshire and Mundie (1966). See Chap. 14 for titration of uronic acids and amino sugars. Equipment and Reagents – Calibrated 150×25 mm glass test tubes; – Water bath;

466

Mineralogical Analysis

– Visible spectrophotometer; – Plastic colorimetric cells, length 1 cm; – Crushed ice; – Concentrated sulphuric acid 18 mol (H2SO4) L–1 (d = 1.84); – 5% phenol solution in water; – 0.2% solution of anthrone in concentrated H2SO4; – Standard solutions for the phenol method: 0, 5, 10, 25, 50 and 100 mg (glucose) L–1 and – Standard solutions for the anthrone method: 0, 5, 10, 15, 20 and 25 mg (glucose) L–1. Procedure for the Phenol Method Put in the test tubes 2 mL of soil hydrolysate (cf. Sect. 13.2.1) or purified soil hydrolysate (cf. Sect. 13.2.2) that has been previously completed to 100 mL in a volumetric flask (a preliminary test of standard additions can be used to check the need for purification). Add 1 mL phenol solution and then rapidly add 5 mL of concentrated sulphuric acid without allowing it to run along the wall of the flask (taking care not to splash). Leave to stand for 10 min, agitate the tubes and place them in the water bath at 25–30°C for 20 min, then cool under running water. Read absorbance at 485 nm (490 nm for hexose sugars, 480 nm for pentose sugars and uronic acids). The colour remains stable for several hours. It is sometimes necessary to homogenize the solutions just before colorimetric reading. Proceed in the same way for each point of the calibration range. Procedure for the anthrone method Bring the hydrolyzed solution (cf. Sects. 13.2.1 or 13.2.2) to 100 mL and homogenize well. Introduce 5 mL of this solution in a calibrated test tube placed in ice. Proceed in the same way for each calibration point. Slowly add in each tube 10 mL of anthrone solution letting it run down the side of the tube; swirl the tube to mix. Seal with a piece of parafilm and immediately place in a 85°C water bath for 35 min. Cool in an ice-tray and place in a colorimetric cell and read absorbance at at 625 nm. Disposable plastic colorimetric cells (1 cm in length) should be used to limit the number of transfers and washings of the concentrated sulphuric acid solutions.

Non-humic Molecules

467

13.2.4 Titration of Sugars by Gas Chromatography Principle The titration of sugars by gas chromatography is rather difficult to implement for two reasons: – Sugars are too polar to be satisfactorily separated directly on the gas chromatographic columns; it is thus necessary to form derivatives which transform the hydroxyl functional groups into less polar forms; – Gas chromatography is more selective than liquid chromatography; separation of the carbohydrate can result in many peaks representing the isomers of the different molecular configurations and the chromatograms may then be difficult to interpret. Most authors used techniques similar to the one described by Oades et al. (1970). Sodium borohydride was added to the neutralized and purified acid extracts, and this transformed the isomers of sugars into their alditol form. After elimination of the boric acid formed by successive evaporations in acetic acid medium, acetylation of the alditols was performed with acetic anhydride. Alditol acetates were dissolved in methylene chloride for injection into the chromatograph. Oades et al. (1970) used a 2 m column with an interior diameter of 3.5 mm filled with 100–120 mesh GAS Chrom Q impregnated with 5% ECNSS-M. In this way eight major soil sugars were separated in 70 min though with rather poor distinction between rhamnose and fucose. Spiteller (1980) improved this technique. He separated alditol acetates on a non-polar OV1 25 m capillary column in 25 min by detecting traces of glucosamine and galactosamine. Cheshire et al. (1983) separated alditol acetates with a capillary column of 50 m × 0.3 mm impregnated with SILAR 10 CP with a 90 min temperature programme. Dormaar (1984) obtained separation with a duration similar to that used by Spiteller (1980) with a glass capillary tube impregnated with SP2330. Baldock et al. (1987) used the procedure of Spiteller, as did Arshad and Schnitzer (1987), the latter authors with a capillary tube and a stationary phase with trifluoropropyl-methyl which increased the total duration of the analysis. Coelho et al. (1988) used a filled column, and the length of the separation phase was similar to that of Oades et al. (1970). Like Cheshire and Griffiths (1989) and Murayama (1988) used chromatographic separation of alditol acetates but did not specify the column used. The reduction and acetylation operations which follow the purification of the hydrolysates are rather long and this means many samples cannot

468

Mineralogical Analysis

be compared. Blakeney et al. (1983) recommended a method allowing the preparation time to be reduced, but a test performed at the IRD laboratory with this aim in view was unsuccessful (unpublished data). On the other hand, the chromatographic time (Fig. 13.1) was reduced to 12.5 min using a capillary column impregnated with a SP2330 phase similar to that of Dormaar (1984), but made of silica glass instead of borosilicate glass (Pansu 1992). Equipment and Reagents – 5 mL conical cylinder flasks with PTFE joint screw caps (Fig. 13.2); – Pasteur pipettes with 3 mL squeeze bulbs; – Thermostated aluminium heating block (Fig. 13.2); – Nitrogen sweeping for evaporation (Fig. 13.2); – Gas phase chromatograph equipped with a flame ionization detector; – SP2330 silica capillary column (Supelco) 15 m in length and 0.25 mm ID; – Standard carbohydrate solution in 85% methyl alcohol containing 1 mg mL−1 of each carbohydrate: rhamnose, fucose, ribose, arabinose, xylose, mannose, galactose and glucose; – Standard solution containing 1 mg mL–1 myoinositol in 50% methyl alcohol; – Methyl alcohol; – Strontium carbonate; – Sodium borohydride; – Glacial acetic acid; – Anhydrous solution of 10% acetic acid in methanol (dry on anhydrous Na2SO4); – Acetic anhydride and – Chloroform. Preparation of Alditol Acetates (a) In each boiling flask containing the purified hydrolysates, add an exact volume of the myoinositol internal standard solution appropriate for the estimated quantity of sugars (0.5–2 mL). (b) Evaporate to just dry with a rotary evaporator; rinse the boiling flask with 3–4 small fractions of methanol using a Pasteur pipette and transfer the washing solutions to a 5 mL flask with screw cap. (c) Add approximately 10 mg of sodium borohydride and leave to act overnight. (d) Add 0.1 mL of glacial acetic acid, evaporate to dry at 70°C under a nitrogen flow (Fig. 13.2), add 1 mL of anhydrous 10% acetic acid in

Non-humic Molecules

469

the ethyl alcohol solution and evaporate in the same way, repeat this operation five times. (e) Add 1 mL acetic anhydride, stop the flasks and heat at 135°C for 2 h. (f) Cool to 70°C and evaporate to dry under a nitrogen flow. (g) Cool and dissolve in an exact volume of from 0.5 to 2 mL of chloroform depending on the estimated sugar content. The solutions can be stored or injected directly into the chromatograph. Preparation of Standards for Alditol Acetates Add 1 mL standard solution of sugars and 1 mL of internal standard solution in a 5 mL flask with a screw cap and continue as mentioned earlier starting at stage c. Chromatographic Conditions – Silica glass Supelco SP2330 capillary column (or similar) length: 15 m, interior diameter: 0.25 mm. – Carrier gas: 0.7 bar helium. – Splitter injector, leak-flow: 100 mL min–1. – Injection: 1 µL. – Flame ionization detector. – Temperatures: column programmed from 210 to 250°C at 3°C min –1, injector: 300°C, detector: 250°C. Figure 13.1 shows an example of chromatograms obtained on sugars in a standard solution and in a ferrallitic soil from Congo.

470

Mineralogical Analysis

Fig. 13.1. Chromatograms of alditol acetates (conditions described in “Chromatographic conditions”, RH, rhamnose; FU, fucose; RI, ribose; AR, arabinose; XY, xylose; MA, mannose; GA, galactose; GL, glucose; MY, myoinositol internal standard). On the left: standard mixture corresponding to the injection of 0.2 µg of each sugar. On the right: sugars of a strongly desaturated ferrallitic soil of the Niari valley (Congo). Hydrolysis according to the procedure described in “Attack Favouring the Measurement of Pentoses” earlier.

Non-humic Molecules

471

Fig. 13.2 Recommended system for derivatization reactions for GPC. Top, reactions in closed micro-flasks at controlled temperature, bottom, evaporation of excess solvents and reagents under surface nitrogen flow

472

Mineralogical Analysis

13.2.5 Quantification of Total Lipids Principle Total lipids are measured gravimetrically after extraction with organic solvents in a Soxhlet extractor. Most lipidic compounds of soils are not very polar, but some are slightly more so (Amblès et al. 1990). The choice of the polarity of the extraction solvent is thus significant, and preliminary tests should be performed. Fahd-Rachid (1993) used chloroform. A mixture of petroleum ether and ethyl acetate (3:1, v:v) is also often recommended and its toxicity is lower than chlorinated solvents. However, this mixture was considered to be less effective than chloroform for the extraction of complex lipids (Jambu et al. 1987). Solvents do not extract all the lipids since some are associated with humic compounds, clays and various minerals. An acid treatment enables release of the bound lipids which then become accessible for a second Soxhlet extraction. Hydrochloric acid can be used for the acid pretreatment but according to a technique recommended by Wang et al. (1969), it is better to use a mixture of hydrochloric acid and hydrofluoric acid. Hydrochloric acid enables release of lipids bound to cations, and hydrofluoric acid enables release of lipids bound to the organomineral matrix (Fahd-Rachid 1993). Equipment and Reagents – 200 mL Soxhlet extractor with condenser and a 500 mL borosilicate glass round bottom boiling flask with conical ground joints; – Hemispherical electric heating mantle with temperature regulation for use with a 500 mL boiling flask; – Soxhlet extraction cartridges 38 mm in diameter and 150 mm in height; – Vacuum rotary evaporator; – Pasteur pipettes; – 25 mL half cylindrical half conical flasks of the volumetric type (graduation is not necessary) with PTFE joint screw cap; – Glass funnels φ 3 cm; – 250 mL plastic beakers; – Plastic funnels φ approx. 8 cm; – Filter papers for plastic funnels; – HCL–HF aqueous solution at 2.5% HF and 2.5% HCl and – Petroleum ether:ethyl acetate extraction solution (3:1 v:v): mix 750 mL petroleum ether and 250 mL ethyl acetate.

Non-humic Molecules

473

Procedure Free Lipids – Put an exact weight P (100–150 g) of air-dried soil sieved to 2 mm in the extraction cartridge and put the cartridge in the Soxhlet extractor. – Put 200 mL extraction solution and two pumice grains in the 500 mL boiling flask; connect the boiling flask to the Sohxlet extractor. – Connect the condensor to the top of the extractor. – Start heating the boiling flask and regulate the heat of the hemispherical mantle so as to obtain good condensation of the solvent from the condenser to the extractor. – Continue the extraction for 48 h. – Let cool then connect the boiling flask to the rotary evaporator and evaporate until the volume is reduced to a few mL. – At the same time, tare a stopped 25 mL half cylindrical half conical flask (after drying in the drying oven and cooling). – Open the flask and position it under a small funnel stopped with a glass wool plug and half filled with anhydrous sodium sulphate. – Using a Pasteur pipette, decant the residue of evaporation into the small funnel and collect the dried extract in the 25 mL flask. – Rinse the evaporation flask and the funnel several times with a few Pasteur pipettes of the extraction mixture to fill approximately 20 mL of the 25 mL flask. – Evaporate the solvent from the flask under gaseous nitrogen flow (see Fig. 13.2) or under vacuum in the rotary evaporator by means of a ground/screwed connection. – Stop the flask, let cool and weigh, by deduction of the tare one obtains the weight of free lipids P1. – Rate of free lipid = 100 P1/P. Bound Lipids – Place the residue of extraction of the free lipids in a 250 mL plastic beaker and add the HCL–HF solution at a rate of 150 mL for a test sample of 100 g soil. – Leave in contact for 48 h agitating from time to time. – Filter on filter paper in plastic funnels. – Wash the residue with water until pH 5; decant it to a shallow cup or glass beaker cover and leave to dry on the lab table or in a desiccator. – Extract the lipids using same procedure as for free lipids (cf. “Free Lipids” earlier).

474

Mineralogical Analysis

Total Lipids It is possible to extract the total lipids directly by an attack of the initial sample with the HCL–HF mixture followed by extraction in the Sohxlet. Fahd-Rachid (1993) found excellent agreement between the total extracted lipids and those calculated by the sum “free lipids + bound lipids”. 13.2.6 Quantification of the Water-Soluble Organics
Table 13.1.1. Cold water soluble (CWS) and hot water soluble (HWS) compounds obtained on cultivated soils of Boigneville, France (Pansu, unpublished data, see extraction conditions in the text) weight % of CWS 0.026 0.033 0.004 0.021 0.025 CWS C:N ratio 2.1 2.5 2.1 1.4 1.8 % CWSC/total soil C 0.75 1.12 0.30 0.57 0.75 C% in HWS 8.3 22.3 21.5 25.1 22.4 HWS C:N ratio % HWSC/total soil C 7.1 10.7 5.6 6.0 8.5

No. A B D E F

C % in CWS 28.8 33.8 75.0 27.1 30.1

weight % of HWS 0.85 0.48 0.26 0.24 0.38

5.5

For the study of the spatial and temporal dynamics of organic matter, the water-soluble organic fractions of the soil have to be taken into account. These fractions comprise organic acids, simple sugars or light polysaccharides and nitrogenous compounds. These different compounds can have a significant influence on the structural stability and fertility of the soils. Two types of compounds can be distinguished: – Compounds that are soluble in cold water: these are obtained by agitating the soil with water using different procedures, for example, shake 10 g of soil in 200 mL water or 2 h on a rotary shaker, leave in contact overnight, sake again for 2 h and centrifuge at 14,000 g. – Compounds that are soluble in hot water: several procedures can be used, for example, boil 4 g of soil at reflux for 16 h with 200 mL water in the presence of three glass balls, cool and centrifuge at 14,000 g. Kouakoua et al. (1997) showed that the extraction of water-soluble compounds of ferrallitic soils from Congo increased continuously with the length of extraction both in a drying oven and in an autoclave. According to Leinweber et al. (1995), this fraction is mainly made up of nitrogenous and carbohydrate compounds. Table 13.1 lists proportions of these fractions expressed in mass, carbon and nitrogen contents for five cultivated soils in temperate zones.

Non-humic Molecules

475

In waterlogged or poorly aerated soils, the aqueous extracts contain organic acids of low molecular weight (e.g. lactic, pyruvic or acetic acid) under anaerobiosis conditions (Küsel and Drake 1999) which can be titrated in aqueous mediums by gas–solid chromatography (standard column of Porapak or chromosorb 101, or similar) or by ionic chromatography. The aqueous extracts contains also inorganic water soluble compounds (cf. Chap. 18).

13.3 Complementary Techniques
13.3.1 Determination of soil Carbohydrates by Gas Chromatography If the alditol acetate technique (cf. Sect. 13.2.4) was the most widely used for the measurement of soil carbohydrates, other techniques have also been proposed: Morgenlie (1975) described separation of neutral sugars in the form of their O-isopropylidene derivatives. Preparation of the derivatives was performed with 1% sulphuric acid in acetone reagent, and separation required 35–40 min on columns of the XE60 or OV225 type. Traitler et al. (1984) separated trimethylsilyl derivatives of sugars on short apolar columns. Cowie and Hedges (1984) also separated trimethylsilyl derivatives from sugars from hydrolysates in plankton, sediment and wood. As a preliminary treatment, these authors brought free sugars to their mutarotation equilibrium in the presence of lithium perchlorate to get round the problem of multiple peaks. Larre-Larrouy and Feller (1997) and Larre-Larrouy et al. (2003) analyzed neutral sugars in soil at the same time as uronic acids and hexosamines in the form of their trimethylsilyl derivatives, each sugar being calculated by the sum of the surfaces of its different isomer forms.

13.3.2 Carbohydrates by Liquid Chromatography The techniques used for the titration of sugars with liquid chromatography were unsatisfactory for many years. Cheshire et al. (1969) separated eight neutral sugars by ion exchange chromatography with a pH gradient, but separation took 14 h. At the column exit, sugars were analyzed by colorimetry after reaction with the orcinol in the sulphuric acid reagent. Hydrazide of p-hydroxybenzoic acid is a better reagent for alkaline eluates; without acidification it gives an yellow colour with carbohydrates.

476

Mineralogical Analysis

Hamada and Ono (1984) analyzed sugars on soils of volcanic ash by high performance liquid chromatography (HPLC) with an anion column and detection by fluorescence spectroscopy after reaction with ethanolamine. Separation required 70 min but arabinose, fructose and fucose were eluted under the same peak, as were rhamnose and ribose. Pluijmen (1987) analyzed sugars of different plants by HPLC on a SUGAR PAK TM column (Waters Associates) with a water (or acetonitrile–water) mobile phase, refractometric detection and an anion and cation precolumn. But this author was especially concerned with glucose and fructose and did not provide chromatograms. Reim and Van Effen (1986) used a technique that appeared to be more promising. They proposed a new detector that was more sensitive than the refractometer and which, in addition, did not require preliminary derivatization reactions as do spectrometric methods. Using an ion exchange column, they simultaneously separated simple sugars and low molecular weight oligomers, but they did not provide a chromatogram of the eight main soil sugars. Angers et al. (1988) separated sugars from hydrolysates of soils on an aminex HPX-87P column (BIO-RAD labs); but they titrated only five sugars and the detection limit was rather poor. Martens and Frankenberger (1990) separated ten soil sugars by anion exchange chromatography using the HPAC-PAD system (Dionex, Sunnyvale, CA, USA) with the following chromatographic conditions (Fig.13.3): – 200 µL injection loops; – CarboPac PA guard column (25 × 3 mm); – Chromatographic column (250 × 4 mm) filled with a pellicular anion exchange resin: CarboPac PA1; – Flow of eluent: 0.8 mL min –1, ambient temperature; eluent a: purified inorganic water (18 Mohm), eluent b: 50 mmol (NaOH) L–1 + 1.5 mmol (CH3COONa) L–1 aqueous solution, 93% eluent a and 7% eluent b for 15 min, gradient up to 100% b in 25 min, idem the mobile phase has to be degassed to prevent absorption of CO2 and the production of carbonates which can move ions and reduce the retention time; – Detection by pulsed amperometry three times with a gold electrode: E1:0.1 V, t1: 300 ms, oxidation of CHOH groups, E2: 0.6 V, t2: 120 ms, displacement of the reaction products, E3:–0.8 V, t3: 300 ms, cleaning of the electrode at negative potential, response time: 1.

Non-humic Molecules

477

The study of Martens and Frankenberger (1990) showed that the HPAC-PAD system had several advantages over another classical HPLC system with detection by refractive index: it produced more precise results, was twice as sensitive (pmole) and had better resolution. The preparation of the soil samples was also simpler in the HPAC-PAD system. After acid attack (cf. Sect. 13.2.1), the samples were treated with 1 mL 0.1 mol (EDTA) L–1 solution then brought to pH 4 by adding 5 mol (NaOH) L–1 solution and centrifuged at 10,000 g. The coloured materials were then removed by filtration on a solid phase extraction column (SPE) (SupelcoTM Bellefonte, Pa, USA) comprising 3 mL of strong cation exchange resin (three propylsulfonic acid, H+ form) and 3 mL of strong anion exchange resin (quaternary propylammonium 3, Clform). The extracts were also filtered on 0.22 µm GS filters (Millipore, Bedford, MA, USA).

Fig. 13.3. Standard sugar separation by high performance anion exchange chromatography (Martens and Frankenberger 1990); Ino, inositol; Rib, ribitol; Fu, fucose; Rh, rhamnose; Ga, galactose; Gl, glucose; Xy, xylose; Ma, mannose; Ri, ribose; Lac, lactose

Bernal et al. (1996) proposed a chromatographic technique similar to that of Martens and Frankenberger (1990) for sugar titration in coffee and wine.

478

Mineralogical Analysis

13.3.3 Fractionation and Study of the Soil Lipid Fraction Lipid Fractionation There are many lipid fractionation techniques and the reader is advised to consult relevant handbooks that describe biochemical lipidology techniques, e.g. Kates (1975). Fractionation can be divided into two stages: – Separation of the different classes of lipids and – Measurement of the individual components of each class or of the total fraction. The exact procedure to apply depends on the lipid concerned. Most lipids of microbial and animal origin contain from 60 to 85 % of phosphatides and glycolipides, the remainder being neutral or not polar lipids (glycerides, sterols, hydrocarbons, pigments). Lipids of plant origin contain a larger proportion of neutral lipids and a smaller proportion of phosphatides. Two main types of techniques are used to separate the classes of lipids: – Fractionation by solvents and – Liquid phase chromatography on column, on paper or on thin layer. The characterization and quantification of lipid components require the chemical breakdown of complex lipids followed by separation techniques such as gas chromatography, generally with a flame ionization detector, possibly coupled with a mass spectrometer for the identification of the molecules. Fractionation by Solvents Precipitation by acetone is the simplest method for separating phosphatides and neutral lipids. It is based on the insolubility in cold acetone (0°C) of most phosphatides (particularly of acid phosphatides in salt forms), whereas neutral lipids are soluble. This procedure is well suited for lipids of animal and microbial origin, but is less effective in the presence of high proportions of neutral lipids like most lipids of plant origin (Kates 1975). The techniques for separation of nonmiscible pairs of solvents (polar and non-polar) are also useful, especially for lipids rich in glycerides. As the first stage of the soil lipid fractionation, Stevenson (1982) recommended the separation of chloroform and an aqueous soda solution. The aqueous phase was then acidified and extracted with ether to recover the free fatty acids. The chloroform phase contains the other lipids which were then separated by chromatography. Fractionations by Liquid Phase Chromatography Many methods have been proposed. For preliminary fraction-ation of the total lipidic extract of the soil, Jambu et al. (1991, 1993) and Amblès et al. (1989, 1990) used the technique recommended by McCarthy and

Non-humic Molecules

479

Duthie (1962). Lipids were separated on columns of potassic silica (silicic acid treated with potash in isopropanol) in three fractions: neutral (the first elution with ethyl ether), acid (elution with 2% formic acid in ether solution) and polar. Zelles and Bai (1993) used a similar technique with SPE. It was then possible to separate the neutral fraction of lipids on silica columns (Kiesselgel 60, Merck, Jambu et al. 1993) or Florisil treated or not with acid (Kates 1975). Elution was carried out initially by hexane or petroleum ether to separate hydrocarbons then by mixtures of increasing polarity to split the other classes of lipids. Mixtures with an increasing proportion of ethyl ether in petroleum ether were used most frequently enabling successive separation of esters (sterilic, methyl esters), ketones, triglycerides, diglycerides, monoglycerides, free alcohols and sterols. Gas–Liquid Chromatography Techniques These techniques are often used for fractionation, characterization and quantification of lipid components. They can be applied directly to certain fractions like hydrocarbons or after breakdown of the heavy lipidic compounds into fatty acids and unsaponifiable products (cf. “Fractionation of Fatty Acids and Unsaponifiables”). Fractionation of Fatty Acids and Unsaponifiables Principle A saponification reaction enables breakdown of heavy lipidic substances to obtain fatty acids, and other saponification products (like glycerol and sterols) with unsaponifiable compounds. The fatty acids are then methylated, separated, and quantified by Gas–Liquid Chromatography (GLC), the unsaponifiables are also titrated by GLC after silylation of the hydroxyl groups. The technique can be used for total lipidic extracts or for lipid groups that have already been separated as described in “Lipid Fractination” earlier. Equipment and Reagents – F1 flasks: 25 mL half cylindrical half conical borosilicate glass flasks, (volumetric flasks without graduation, cf. “Equipment and Reagents” earlier under Sect.13.2.5 and Fig. 13.4) with PTFE screw cap. – F2 flasks: 25 mL half cylindrical half conical flasks with PTFE screw cap. – F3 flasks: 10 mL half cylindrical half conical borosilicate glass flasks, (volumetric flask without graduation) with PTFE screw cap.

480

Mineralogical Analysis

– F4: half cylindrical half conical borosilicate glass volumetric tubes with PTFE screw caps. – Funnels with a diameter of 3 cm. – Pasteur pipettes. – 0.1 mg precision balance. – Gas phase chromatograph with splitter or “split-splitless” injector for capillary column and detection by flame ionization detector. – Anhydrous petroleum ether. – Anhydrous methanol (stored on anhydrous sodium sulphate). – 0.3 mol (NaOH) L–1 solution in methanol: dissolve 1.2 g of soda in 100 mL anhydrous methanol. – 3% H2SO4 methanolic solution: add 3 mL of concentrated sulphuric acid in a 100 mL volumetric flask containing 70 mL anhydrous methanol, agitate and cool, adjust to 100 mL and stop well. – 60% methanol in water. – Anhydrous sodium sulphate. Procedure Saponification – Add 4 mL of petroleum ether and 2 mL methanol in the 25 mL F1 flasks containing the free or bound lipidic extracts (cf. “Procedure” under Sect. 13.2.5), agitate for 1 h. – Add 8 mL of 0.3 mol (NaOH) L–1 solution in methanol. – Stop the flasks hermetically and heat at 100°C for 2h30 in a thermostated aluminium heating block (Fig. 13.2). – Cool, open the flasks and add 60% methanol in such a way that the petroleum ether phase is in the upper cylindrical part of the flask. – Using a Pasteur pipette, take the upper phase and decant it in a tared 25 mL conical F2 flask through a small funnel stopped with a glass wool plug and filled with a spatula of anhydrous sodium sulphate (Fig. 13.4). – Add 3 mL of petroleum ether to the first F1 flask; stop and agitate well then remove the upper phase in the same way and add it to the previous phase through the same device; repeat this procedure four times. This phase contains unsaponifiables and saponification products other than fatty acids (F2). – Acidify the first flask with 1 mL of concentrated hydrochloric acid diluted two times, then extract again with petroleum ether as previously described. Decant and collect the upper phase in a tared 10 mL F3 flask; this phase contains the fatty acids.

Non-humic Molecules

481

13.1)

Fig. 13.4. Separation by decantation (on the left) and drying of the recovered products (on the right)

Non Fatty Acid Compounds – Evaporate the contents of the F2 flask to dry under nitrogen flow or with a rotary evaporator. – Stop and weigh the flask to obtain the total quantity. – Add 100 µL of trimethylchlorosilane (TMCS), 300 µL of hexamethyldisilane (HMDS) and 600 µL pyridine, stop and heat at 70°C for 5 h (Fig. 13.2). – Inject 2 µL in the gas phase chromatograph under the following conditions: (a) CPWAX capillary column (or similar) length: 25 m, interior diameter: 0.3 mm, temperature 230°C, (b) carrier gas He 0.5 B, (c) split injector, leak flow 50 mL min–1.

482

Mineralogical Analysis

Fatty Acids – Bring the contents of the 10 mL F3 flasks to dry and weigh to obtain total fatty acids. –Add 3 mL of anhydrous 3% sulphuric acid in the methanol reagent; stop the flasks and heat at 70°C for 5 h (Fig. 13.2). – Let cool and add 2 mL petroleum ether and approximately 3 mL water to recover the ether phase in the cylindrical part of the flask (Fig. 13.4). – Recover the fatty acid methyl esters in 10 mL F4 tubes using the same procedure as for the non-fatty acid lipidic fraction, complete to 10 mL and inject 2 µL in GPC under the following conditions (short acids ≤ C20, Fig. 13.5): (a) 25 m × 0.3 mm silica glass capillary column impregnated with CPWAX phase (or similar), (b) He carrier gas, input pressure 0.6 B, (c) Split injector 180°C, leak-flow 60 mL min –1,
2 °C min (d) Temperature programme 180°C ⎯⎯⎯⎯⎯ 240 °C, → (e) Flame ionisation detector, 240°C.
−1

Fig. 13.5. Fractionation of methyl esters of short fatty acids of a tropical ferruginous soil from Burkina–Faso by gas chromatography (conditions described in the text; M Pansu, unpublished data)

Non-humic Molecules

483

Notes – Some reagents are much faster than MeOH–H2SO4 making it possible to obtain fatty acid methyl esters instantaneously at room temperature (e.g. BF3–CH3OH or diazomethane); the reagent described in “Procedure” under Sect. 13.3.3 earlier was used because it is not expensive and the heating system is suitable for derivatization. – For soils poor in fatty acids, the 10 mL volumetric half cylindrical half conical F4 tubes are very useful for the concentration of the mixture by solvent evaporation under nitrogen flow. – The chromatographic column used to obtain separation of Fig. 13.5 is not suitable for the fractionation of long chain fatty acids of soil (C20– C34); for this purpose a shorter column or a different impregnation phase is required. – The procedure described in “Equipments and Products” later can be simplified for the determination of fatty acids alone; use direct transmethylation on the lipidic fraction with the H2SO4–methanol reagent; extract the fatty acid methyl esters by decantation in petroleum ether after adding water (Fig. 13.4); however, this technique may complicate the reading of the chromatograms due to a risk of peaks of fatty acid methyl esters overlapping peaks of unsaponifiable compounds. 13.3.4 Measurement of Pesticide Residues and Pollutants Principle Like lipids, pesticides and pollutants are soluble in organic solvents. They can be extracted using similar techniques (cf. Sect. 13.2.5), although extraction with the Soxhlet apparatus is not recommended for pesticides. Indeed, the molecules are often degradable and prolonged boiling of the mixture of solvent and extracted products could lead to underestimation of the contents of residues. Due to problems of degradability, the most commonly used methods use cold extraction in solvents, specially micro-methods (Steinwandter 1992) and methods using supercritical fluids (Richter 1992; Table 13.2). If the soil samples have to be stored before extraction, they should be frozen or freeze-dried. After the extraction phase, if lipid content is high, purification may be necessary. Two techniques are available (a) separation using solvents of different polarities and (b) liquid chromatography on a Florisil or alumina column. It is usually necessary to concentrate the soil extracts as much as possible to improve the limit of detection. It is useful to carry out the final

484

Mineralogical Analysis

concentration phase using a gaseous nitrogen flow on the surface of the extract in half cylindrical half conical volumetric tubes (cf. Sect. 13.3.3).
Table 13.2. Some supercritical fluid techniques used for the extraction of soil pesticide residues (Richter 1992) pesticide permethrin, atrazine, deltamethrin, dieldrin, carbofuran, diuron, 2,4 D, methylparathion DDT DDT lindane, aldrin, DDT linuron, diuron extraction conditions Methanol 250 °C, 150 bars, 1 mL min–1, 2h CO2, 40 °C, 100 bars, 0.7 g s−1, 10 min CO2 + 5% methanol or toluene, 40 or 100°C, 0.7 g s–1, 5 min CO2, 138 bars, 15 min CO2 + methanol, ethanol or acetonitrile, 75 to 120°C, 100–400 bars, 2.5–8.5 mL min –1 , 15–180 min CO2 + 2% methanol, 40°C, 223 bars, 6 mL min-1, 2–15 min CO2 to 48°C, 230 bars, 1.7 mL min–1, 30 min CO2 100% and CO2 + 10% methanol, 50–70°C, 150 or 300 bars, 25–60 min CO2 + 5% acetone, 75°C, 400 bars, 60 min

sulfonylurea simazine, atrazine, propazine, terbutylazin, cyanazine organochlorides, organophosphorus DDT, DDE, DDD, lindane, aldrin

The recommended choice for the analysis of the extract is gas chromatography because of (a) its good selectivity, particularly with capillary columns and (b) the large number of selective and very sensitive detectors that are available. The recommended injector for most nonvolatile pesticides is a glass-needle injector which ensures the maximum possible concentration of the solutes during injection. Alternately splitsplitless injector can be used. Other techniques are based on HPLC, generally with UV detection. The diversity of the pesticides and pollutants present in the soil (cf. Sect. 13.1.4 later) has resulted in the development of a very large number of specific analytical procedures which cannot be exhaustively described here. Based on the literature and laboratory practice, a choice of procedures for the extraction, purification and chromatographic analysis of the main families of products is provided later. In the future, these procedures will change as a result of further advances in analytical chemistry and some of the recommendations later will need to be

Non-humic Molecules

485

updated. For example, Gennaro et al. (1996) presented the simultaneous separation of phenoxyacetic acid, triazins and phenylureas herbicide residues in soils by HPLC. Celi et al. (1993) proposed a way of measuring fenoxatrop and ethyl fenoxatrop in a range of soils. Anon (1993) proposed a method for the simultaneous measurements of 27 herbicides in the soil using HPLC. Kiang and Grob (1986) proposed a method for simultaneous measurement of 49 soil pollutants by capillary gas chromatography. Equipment and Products – Back and forth shaker. – Decanting funnels in borosilicate glass with ground Teflon stoppers: 50, 125, 250 mL. – Columns for liquid chromatography, diameter: 1 cm, length: 50 cm, with a PTFE stopcock and a solvent tank. – 10 mL half cylindrical half conical volumetric tubes. – Rotary evaporator with 50, 125, 250 mL boiling flasks. – Regulated aluminium heating block and system of evaporation under flow of nitrogen (Fig. 13.2). Alternatively, a centrifugal evaporator under vacuum, type: speed-vac, can be used. – Gas phase chromatograph with several selective detectors (e.g. electron capture detector for halogenated compounds, thermoionic detector for organophosphorous compounds) and non-selective (e.g. flame ionization or mass spectrometry). The use of a glass needle injector is recommended, or failing that a split-splitless or on column injector (Pansu et al. 2001). – HPLC with UV detector if required; – Certified solvents, free from pesticides and pollutants. Each solvent should be tested by injection in the chromatograph after concentration by evaporation in the rotary evaporator (approximately 200 mL giving 1 mL). The most commonly used solvents are acetone, petroleum ether, hexane, ethyl ether, methanol, acetonitrile, dichloromethane, ethyl acetate. – Different standard pesticides. – Activated florisil (60–80 mesh particle size, 3% of water). – Anhydrous sodium sulphate. – Sodium chloride.

486

Mineralogical Analysis

Extractions Acid Herbicides Jensen and Glass (1990) tested several techniques before recommending cold extraction with ethyl ether in acid medium. When well agitated for a rather long period, this technique is effective even for residues that have been incorporated in the soil for a year. A sufficient volume of water and acid have to be added to reduce the viscosity of the soil sample and ensure good contact with the ether phase. The authors advised three successive extractions with ether and one hour agitation each time. Triazine Herbicides Several different solvents can be used. For extractions in aqueous medium, a mixture of ethyl ether and petroleum ether (v/v) may be appropriate. For soils, we prefer a 1/5 mixture of methylene chloride and ethyl acetate. We recommend three successive extractions with a back and forth shaker on 20 g of sample with respectively, 100, 50 and 50 mL of solvent mixture. Organochlorides Wheeler and Thompson (1990) reviewed the large number of techniques used for the extraction of this type of compound. Methods used for soils are often based on those used in food analysis. A simple way is to agitate the soil sample (50 g) with acetone (100 mL) or an acetone-petroleum ether mixture on a back and forth shaker. With a wet sample, add anhydrous sodium sulphate as desiccant. Organophosphorous Freeze-drying is usually advised for the conservation of soil samples. Several solvents or mixtures of solvents have been used for standard extraction of organophosphorous compounds in the soil: acetone, acetone–water, dichloromethane, ethyl acetate, acetone–hexane, acetone–dichloromethane, methanol–water (Barcelo and Lawrence, 1992). These compounds can be extracted jointly with organochlorides (acetone, acetone–hexane), pyrethrinoids (acetone–hexane), and carbamates (acetone–dichloromethane). Carbamates Carbamates can be titrated with a rather complex technique of multiresidue analysis (Seiber 1990). The following method was tested by Pansu et al. (1981a) with an average extraction yield of 72 ± 8%. Prepare the soil samples by freeze-drying, sieve to 2 mm particle size and divide into sample specimens of 10 g each. Shake the sample specimen with 50 mL methanol and 50 mL water for 6 h on a back-and-forth shaker, add

Non-humic Molecules

487

100 µL of 1 mol (HCl) L–1 solution and filter under vacuum. Wash the residue with the extraction solution and add the washing solutions to the filtrate. Extract the filtrate three times in a 200 mL decanting funnel with respectively, 30, 20 and 20 mL chloroform. Pyrethrinoids In biological substrates, pyrethrinoids are extracted satisfactorily by two successive extractions with hexane in the presence of anhydrous sodium sulphate (Pansu et al. 1981b). For soils, a mixture of hexane with 3% acetone is more efficient. Preparation of the Extracts for Chromatography Principle Most extracts have to be purified before chromatography to eliminate interference with other lipidic compounds in the soil. The use of detectors that are selective for the main families of products (such as organochlorides or organo-phosphorus) makes it possible to limit the number of purifications. However, some of these detectors (electron capture for example) are very sensitive to pollution and the extracts need to be purified so as not to perturb the sensitivity of the detector. The extracts should be concentrated as much as possible to improve detection of ultra-trace residues. Extracts are usually concentrated in vacuum rotary evaporators during the first stage by micro-techniques such as gaseous nitrogen flow (Fig. 13.2) or a speed-vac centrifugal evaporator in the final stage. If too polar solvents are used for extraction (not easily eluted from chromatographic columns) or if they are incompatible with the detection system, the extract may have to be transferred into another solvent before injection. For example, if an electron capture detector is being used, chlorinated solvents have to be completely eliminated before injection. Except for volatile pesticides, this is accomplished by drying the extract several times, each time with dissolution in a non-halogenous solvent. Finally, some polar compounds require derivatization reactions before injection into the chromatograph. Acid Herbicides Purification can be achieved by separation into solvents (Jensen and Glass 1990). The ether extracts (see “Acid Herbicides” in “Extractions” under Sect.13.3.4) are agitated with a sodium bicarbonate aqueous solution. In the aqueous solution, acid herbicides are transformed into salts, while the

488

Mineralogical Analysis

majority of the coloured organic compounds remain in the organic phase. This phase is eliminated, and the acid herbicides are re-extracted with ether after acidification of the aqueous phase and saturation by sodium chloride. If gas chromatography is used for analysis, the extracts usually need to be methylated before injection, as acid herbicides are too polar. Several different reagents can be used in techniques similar to fatty acid methylation (cf. “Procedure” under Sect. 13.3.3 earlier). Diazomethane is a particularly effective methylation agent. Triazines The main lipids can be eliminated by acetonitrile-hexane partition. Concentrate the methylene chloride-ethyl acetate extracts to 2 mL (cf. “Triozine Herbicides”). Add 20 mL acetonitrile saturated with petroleum ether (AN). Add 10 mL petroleum ether saturated with acetonitrile (PE) and agitate. Recover the AN phase. Re-extract PE with 10 mL AN. Discard PE. Add all the AN phases in the decanting funnel, add 120 mL water and 10 mL of a NaCl saturated aqueous solution. Extract triazines twice with 50 mL of the ethyl ether/petroleum–ether v/v mixture. Mix the ether phases, dry on sodium sulphate and concentrate to the exact volume desired (2 mL for example) before injection. Organochlorides First the acetone extract must be transferred in a hexane phase. Put the acetone extract in a decanting funnel (cf. “Organochlorides” earlier in “Extractions” under Sect. 13.3.4) with four times its volume of water and 25 mL hexane. Swirl gently and decant the organic phase. Extract the aqueous phase again with 25 mL hexane, mix the hexanic extracts, dry on anhydrous sodium sulphate and concentrate to 5 or 10 mL. Purify the extracts on a column filled with activated Florisil (3% water). Weigh 5 g of Florisil and place it in a column with a diameter of 1 cm stopped with a glass wool plug and partially filled with petroleum ether. Add 2–3 cm of anhydrous sodium sulphate, rinse with 50 mL hexane until the liquid comes to the top the sodium sulphate phase. Deposit the hexanic extract on the column with a Pasteur pipette. Regulate the flow to approximately 15–20 drops per minute with the PTFE stopcock. When the level of the liquid comes to the top of the sodium sulphate phase, add a little hexane while rinsing with the pipette; repeat this procedure once. Elute with exactly the volume of hexane needed to obtain PCB, i.e. approximately 20–30 mL. This E1 eluate may also contain hexachlorobenzene (HCB) and dichlorodiphenylethane (DDE), one of the breakdown products of DDT. The exact quantity of solvent for the E1 eluate must first be determined under the same

Non-humic Molecules

489

conditions with standard mixtures. For PCB, these standards are made of sets of products providing typical chromatograms (e.g. dp5 and dp6 mixtures). After collection of the E1 eluate, elution should be continued with the 10% of dichloromethane in the petroleum ether (or hexane) mixture until a sufficient volume has been obtained for collection of the other organochlorinated compounds (E2 eluate). The exact volume of E2 should first be determined with one or more standard compounds, dieldrin for example. The E2 eluate has to be